WORKING PAPERS

THE EFFECT OF GOVERNMENT POLICY CHANGES ON THE SUPPLY OF

PHYSICIANS EXPANSION OF A COMPETITIVE FRINGE

Monica Noether

WORKING PAPER NO 118

September 1984

FfC Bureau of Economics working papers are preliminary materials circulated to stimulate discussion and critical comment All data contained in them are in the public domain This includes information obtained by the Commission which has become part of public record The analyses and conclusions set forth are those of the authors and do not necessarily reflect the views of other members of the Bureau of Economics other Commission staff or the Commission itself Upon request single copies of the paper will be provided References in publications to FTC Bureau of Economics working papers by FTC economists (other than acknowledgement by a Titer that he has access to such unpublished materials) should be cleared with the author to protect the tentative character of these papers

BUREAU OF ECONOMICS FEDERAL TRADE COMMISSION

WASHINGTON DC 20580

THE EFFECT OF GOVERN MENT POL ICY CHANGES ON THE SU PPLY OF PHYSIC IANS EXPANSION OF A COMPETITIVE FR INGE

Monica Noether

July 1984 Revised September 1984

1 INTR ODUCTION AND BACKGROUND

The American Medical Association (AMA ) has often been

credited wit h effectively restricting the supply of physicians

permitted to practice in the United States for the better part of

the twentieth century 1 Yet since 1965 the active physician

population ratio has risen from about 14 to almost 2 ( per 1000

This paper is based on part of my dissertation The Growing Supply of Physicians Are the Entry Barriers Breaking ( University of Chicago 1983 ) I owe much to my chairman Peter

Pashigian who was always generous wit h his insig hts and time I would also like to thank the other mem bers of my committee John Abowd Sam Peltzman Dennis Carlton Jack Gould and Jody Sindelar for their many valuable suggestions Helpful comments were also received from colleagues at both the University of Chicago and the Federal Trade Commission Financial assis tance was pr ovided by the HB Ear hart Foundation while I was a student

1 Two theories exist to explain the AMAs motivation for restricting physician supply One cr edited primarily to Milton Friedman Occupational Licensure in Capitalism and Freedom ( 1962 ) at 137 and Reuben Kessel Price Discrimination in

Medicine 1 Law and Economics 20 ( 1958 ) views the AMA as a profi maximizing cartel of physicians The other espoused by Kenneth Arrow Uncertainty and the Wel fare Economics of Medical Care 53 American Economic Review 941 ( 1963 ) and Keith Leffler Physician Licensure Competition and Monopoly in Ame rican Medicine 21 J ournal of Law and Economics 165 ( 1978 ) suggests that the AMA exists primarily to assure a certain standard of quality acr oss p hysicians Both stories im ply that the AMA has the market power to restrict the supply of physicians effectively See Monica Noether The Growing Supply of Physicians Has the Market Become More Competitive ( FTC wor king paper 1984 ) for a lengthier discussion of the tw o hypot heses

population) in 1981 after remaining fairly stable since the end

o f World War II The number of Foreign Medical Graduates (FMG)

ac hieving initial US licensure annually more than quadr upled

between 1965 and 19722 Similarly US medical school output

doubled between 1965 and 1980 Can all of this growth in the

number of physicians practicing in the us be explained by an

expansion of demand due to increased government expenditures for

medical care (eg Medicare and Medicaid) as well as hig her

consumer incomes Or is it also tr ue that the AMA has become

less effective at maintaining the entry barriers necessary to

restrict the supply of physicians permitted to practice in the

United States Evidence exists to suggest that the ret urn to

p hysician training has decreased substantially over the la st

decade Figure 1 portrays the ratio of physician income to

opportunity cost3 over the last 37 years While some of that

decline is undoubtedly attributable to a grad ual return to

equilibrium following a rapid and la rge increase in demand in the

1960s the downward trend has continued long enoug h to arouse

suspicions of permanent changes in market structure Moreover

the popular and medical trade press provides numerous indications

2 FMG entry has fallen in the last eig ht years but is still double its 1965 level

3 Opportunity cost is measured wit h college graduate earnings adjusted for additional training time as the benc hmark Its derivation is described in detail in Noether supra note 1

- 2shy

l

f I

-middot ---------------0----

145

1 bullbull o

I J5

1 s

llO

I _ _

_ _ ___ ___ _ _ __ __ _ _ _ _ ampamp n C 12 - a ao

F I GURE l

Phy s i c i a n I ncome - Oppo r tu n i ty Cos t Rat i o 1 9 4 6- 1 9 8 2

lNCrOC 50

I I

I

lO

I I I

bull 15

110

bullamp bulla so sz sbull bull

3 --

--------------------- - ----

that physicians themselves feel that they are operating in a more

competitive envir onment4

AMA authority over the medical sc hool accreditation process

combined wit h mandatory state licensure of individual physicians

has been cited as an effective restriction of the physician

supplyS The AMA has not only controlled the supply of US

medical sc hool spaces by making it di fficult for new sc hools to

enter and by preventing expansion of existing sc hools6 b ut has

also worked to assure that state licensing statutes require

graduation from AMA-accredited sc hools Foreign entry has al so

been curtailed by restrictive licensing laws as well as a strict

federal immigration policy However some of these policies have

been modified in recent years Have such changes led to

increased entry into the medical profession

Despite AMA pr otestations that the supply of US grad uates

was adequate public concern about a physician shortage

developed during the 1 950s A fter several failures in the mid shy

1960s bot h federal and state governments ena cted funding

programs to encourage expansion of existing and creation of ne w

4 See for example Coping with the Approaching Doctor Glut New England Journal of Medicine (1982) cited in AMA to Help MDs Adjust to Market American Medical News 12 November 1982 at 1 and Doctors are Entering a Brave New World of Competition B usiness Week 16 July 1984 at 56-61

5 See for example Kessel supra note 1 Friedman supra note 1 and D R Hyde P Wol ff A Gross and E L Hof fman The American Medical Association Power Purpose and Politics in Organized Medicine 6 3 Yale Law Journal 938 (1954)

6 Schools are accredited for a given size level

- 4shy

medical schools 7 At ab out the same time the federal immigrashy

tion law was eased and thr oughout the 1960s state licensing laws

were relaxed It appears that policies designed to inhibit

growth in the supply of physicians practicing in the United

States have been weakened If these programs were effective

then we should be able to explain at least part of the recent

expansion of the physician stock by the re gulatory and funding

c hanges that have occurred Such evidence would suggest that AMA

authority has dimini shed 8

7 The AMAs position on public funding of medical scho ols changed during the 1 950s from one of total opposition to acknowshyledgment that a one-time grant to support construction prefershyably donated by the private sector would be beneficial During this decade both medical sc hools whic h ca me to rely increasingly on federal research funding and hospit als whose expansion increased demand for medical residents favored strong federal support of medical schools This pressure coupled with that from the general public for increased physician supp ly as well as rising physician incomes during the 1 950s whic h signalled increased demand for medical ser vices perhaps led to the AMAs approval by 1960 o f a controlled expansion of medical schools The funding programs subsequently enacted at both federal and state le vels however far exceeded the AMAs desires For a detailed acco unt of the AMAs ef f orts to thwart the enactment of federal programs to aid medical schools see Elton Rayack Professional Power and American Medicine The Economics of the American Medical Association (1967)

8 Tho mas Hall and Cotton Lindsay Medical Schools Producers of W hat Sellers to W hom 23 Journal of Law and Economics 55 ( 1980) at 58 and 78 cite the gr owth in us licensing of FMGs as

well as medical sc hools positive enrollment response to donashytions particularly federal as evidence that the medical professhys ion possesses no monopoly power M ost notably the pr ofess ion [organized me dicine] has been spectacuarly unsuccessful in preventing expansion of su pply through immigration of foreignshytrained physician s Have me dical schools co nspired with organized medicine to restrict the supply of practitioners Our evidence suggests that they have not Medical sc hool output is

(f ootnote co ntinued)

-5shy

In this paper we model the stock of US practicing physishy

cians as composed of a dominant firm of AMA-sanctioned physicians

and a competitive fringe whose entry is imper fectly restricted

C hanges in the degree of competition in the physician market are

measured by estimating the gr owth of the competitive fringe

component due to the relaxation of vari ous restrictions A s

noted above until the advent of large -scale federal involvement

in medicine in the 1960s it is possible to view us physicians

as a dominant firm that maintained entry barriers pr om ulgated by

the AMA and enforced by federal and state governments su ffishy

ciently hig h to prevent most competitive fringe physicians from

entering The tremendous growth in the physician -population

ratio in the la st 15 yea rs however can per haps be explained by

the increased entry of the competitive fringe due to eroding

entry barriers

As suggested above two sources of a growing competitive

fringe seem likely First de spite AMA ef for ts substantial

numbers of foreign medical gr aduates have been licensed in the

United States part icularly in the last 15 years since many

restrictive laws have been relaxed They certainly pr ovide a

product similar en ough to be considered a substitute Second

the number of us me dical students gr ad uating ann ually has risen

(footnote continues)

positively related to demand by both donors and applicants Interestingly they never seem to con sider the possibility that the AMAs power may have weakened over time

-6shy

-7-

sharply While this may be explained solely by a growth in

demand it is also possible that the AMAs control over the

number of medical school spaces weakened lea ding to the cr eation

o f a domestic competitive fringe The Liaison Committee on

Hedical Education established in 1942 by the AMA and the

American Association of Medical Colleges9 now contains represen shy

tatives for the public and federal government in addition to its

founding organizations While other fact ors such as state

funding programs are also im portant the federal subsidies

granted to me dical schools beginning in the mid-sixties for both

new construction and expansion and as a direct incentive to

expand enrollment (per student capitation grants ) may be viewed

as a summary mea sure of forces weakening the AMAs control over

US me dical schools That is if the AMA as discussed earlier

provided one force which inhibited medical schools from expandshy

ing then the federal government may have supplied an opposing

influence

Actual entry of the dominant firm (DFNL ) in any given year t

is posited to equal a proportion y of the di fference between

the anticipated optimal stock in year t MDSt and the sum of

year t-ls anticipated actual stock MDSt-lbull and the anticipated

fringe entry in year t CNLt pl us a suf ficient number to

replace fully the anticipated depreciation of last years

stock10 That is

Previ ously each organization undertook se parate school surveys

lO The rate of depreciation is assumed to be constant across time and known

9

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

THE EFFECT OF GOVERN MENT POL ICY CHANGES ON THE SU PPLY OF PHYSIC IANS EXPANSION OF A COMPETITIVE FR INGE

Monica Noether

July 1984 Revised September 1984

1 INTR ODUCTION AND BACKGROUND

The American Medical Association (AMA ) has often been

credited wit h effectively restricting the supply of physicians

permitted to practice in the United States for the better part of

the twentieth century 1 Yet since 1965 the active physician

population ratio has risen from about 14 to almost 2 ( per 1000

This paper is based on part of my dissertation The Growing Supply of Physicians Are the Entry Barriers Breaking ( University of Chicago 1983 ) I owe much to my chairman Peter

Pashigian who was always generous wit h his insig hts and time I would also like to thank the other mem bers of my committee John Abowd Sam Peltzman Dennis Carlton Jack Gould and Jody Sindelar for their many valuable suggestions Helpful comments were also received from colleagues at both the University of Chicago and the Federal Trade Commission Financial assis tance was pr ovided by the HB Ear hart Foundation while I was a student

1 Two theories exist to explain the AMAs motivation for restricting physician supply One cr edited primarily to Milton Friedman Occupational Licensure in Capitalism and Freedom ( 1962 ) at 137 and Reuben Kessel Price Discrimination in

Medicine 1 Law and Economics 20 ( 1958 ) views the AMA as a profi maximizing cartel of physicians The other espoused by Kenneth Arrow Uncertainty and the Wel fare Economics of Medical Care 53 American Economic Review 941 ( 1963 ) and Keith Leffler Physician Licensure Competition and Monopoly in Ame rican Medicine 21 J ournal of Law and Economics 165 ( 1978 ) suggests that the AMA exists primarily to assure a certain standard of quality acr oss p hysicians Both stories im ply that the AMA has the market power to restrict the supply of physicians effectively See Monica Noether The Growing Supply of Physicians Has the Market Become More Competitive ( FTC wor king paper 1984 ) for a lengthier discussion of the tw o hypot heses

population) in 1981 after remaining fairly stable since the end

o f World War II The number of Foreign Medical Graduates (FMG)

ac hieving initial US licensure annually more than quadr upled

between 1965 and 19722 Similarly US medical school output

doubled between 1965 and 1980 Can all of this growth in the

number of physicians practicing in the us be explained by an

expansion of demand due to increased government expenditures for

medical care (eg Medicare and Medicaid) as well as hig her

consumer incomes Or is it also tr ue that the AMA has become

less effective at maintaining the entry barriers necessary to

restrict the supply of physicians permitted to practice in the

United States Evidence exists to suggest that the ret urn to

p hysician training has decreased substantially over the la st

decade Figure 1 portrays the ratio of physician income to

opportunity cost3 over the last 37 years While some of that

decline is undoubtedly attributable to a grad ual return to

equilibrium following a rapid and la rge increase in demand in the

1960s the downward trend has continued long enoug h to arouse

suspicions of permanent changes in market structure Moreover

the popular and medical trade press provides numerous indications

2 FMG entry has fallen in the last eig ht years but is still double its 1965 level

3 Opportunity cost is measured wit h college graduate earnings adjusted for additional training time as the benc hmark Its derivation is described in detail in Noether supra note 1

- 2shy

l

f I

-middot ---------------0----

145

1 bullbull o

I J5

1 s

llO

I _ _

_ _ ___ ___ _ _ __ __ _ _ _ _ ampamp n C 12 - a ao

F I GURE l

Phy s i c i a n I ncome - Oppo r tu n i ty Cos t Rat i o 1 9 4 6- 1 9 8 2

lNCrOC 50

I I

I

lO

I I I

bull 15

110

bullamp bulla so sz sbull bull

3 --

--------------------- - ----

that physicians themselves feel that they are operating in a more

competitive envir onment4

AMA authority over the medical sc hool accreditation process

combined wit h mandatory state licensure of individual physicians

has been cited as an effective restriction of the physician

supplyS The AMA has not only controlled the supply of US

medical sc hool spaces by making it di fficult for new sc hools to

enter and by preventing expansion of existing sc hools6 b ut has

also worked to assure that state licensing statutes require

graduation from AMA-accredited sc hools Foreign entry has al so

been curtailed by restrictive licensing laws as well as a strict

federal immigration policy However some of these policies have

been modified in recent years Have such changes led to

increased entry into the medical profession

Despite AMA pr otestations that the supply of US grad uates

was adequate public concern about a physician shortage

developed during the 1 950s A fter several failures in the mid shy

1960s bot h federal and state governments ena cted funding

programs to encourage expansion of existing and creation of ne w

4 See for example Coping with the Approaching Doctor Glut New England Journal of Medicine (1982) cited in AMA to Help MDs Adjust to Market American Medical News 12 November 1982 at 1 and Doctors are Entering a Brave New World of Competition B usiness Week 16 July 1984 at 56-61

5 See for example Kessel supra note 1 Friedman supra note 1 and D R Hyde P Wol ff A Gross and E L Hof fman The American Medical Association Power Purpose and Politics in Organized Medicine 6 3 Yale Law Journal 938 (1954)

6 Schools are accredited for a given size level

- 4shy

medical schools 7 At ab out the same time the federal immigrashy

tion law was eased and thr oughout the 1960s state licensing laws

were relaxed It appears that policies designed to inhibit

growth in the supply of physicians practicing in the United

States have been weakened If these programs were effective

then we should be able to explain at least part of the recent

expansion of the physician stock by the re gulatory and funding

c hanges that have occurred Such evidence would suggest that AMA

authority has dimini shed 8

7 The AMAs position on public funding of medical scho ols changed during the 1 950s from one of total opposition to acknowshyledgment that a one-time grant to support construction prefershyably donated by the private sector would be beneficial During this decade both medical sc hools whic h ca me to rely increasingly on federal research funding and hospit als whose expansion increased demand for medical residents favored strong federal support of medical schools This pressure coupled with that from the general public for increased physician supp ly as well as rising physician incomes during the 1 950s whic h signalled increased demand for medical ser vices perhaps led to the AMAs approval by 1960 o f a controlled expansion of medical schools The funding programs subsequently enacted at both federal and state le vels however far exceeded the AMAs desires For a detailed acco unt of the AMAs ef f orts to thwart the enactment of federal programs to aid medical schools see Elton Rayack Professional Power and American Medicine The Economics of the American Medical Association (1967)

8 Tho mas Hall and Cotton Lindsay Medical Schools Producers of W hat Sellers to W hom 23 Journal of Law and Economics 55 ( 1980) at 58 and 78 cite the gr owth in us licensing of FMGs as

well as medical sc hools positive enrollment response to donashytions particularly federal as evidence that the medical professhys ion possesses no monopoly power M ost notably the pr ofess ion [organized me dicine] has been spectacuarly unsuccessful in preventing expansion of su pply through immigration of foreignshytrained physician s Have me dical schools co nspired with organized medicine to restrict the supply of practitioners Our evidence suggests that they have not Medical sc hool output is

(f ootnote co ntinued)

-5shy

In this paper we model the stock of US practicing physishy

cians as composed of a dominant firm of AMA-sanctioned physicians

and a competitive fringe whose entry is imper fectly restricted

C hanges in the degree of competition in the physician market are

measured by estimating the gr owth of the competitive fringe

component due to the relaxation of vari ous restrictions A s

noted above until the advent of large -scale federal involvement

in medicine in the 1960s it is possible to view us physicians

as a dominant firm that maintained entry barriers pr om ulgated by

the AMA and enforced by federal and state governments su ffishy

ciently hig h to prevent most competitive fringe physicians from

entering The tremendous growth in the physician -population

ratio in the la st 15 yea rs however can per haps be explained by

the increased entry of the competitive fringe due to eroding

entry barriers

As suggested above two sources of a growing competitive

fringe seem likely First de spite AMA ef for ts substantial

numbers of foreign medical gr aduates have been licensed in the

United States part icularly in the last 15 years since many

restrictive laws have been relaxed They certainly pr ovide a

product similar en ough to be considered a substitute Second

the number of us me dical students gr ad uating ann ually has risen

(footnote continues)

positively related to demand by both donors and applicants Interestingly they never seem to con sider the possibility that the AMAs power may have weakened over time

-6shy

-7-

sharply While this may be explained solely by a growth in

demand it is also possible that the AMAs control over the

number of medical school spaces weakened lea ding to the cr eation

o f a domestic competitive fringe The Liaison Committee on

Hedical Education established in 1942 by the AMA and the

American Association of Medical Colleges9 now contains represen shy

tatives for the public and federal government in addition to its

founding organizations While other fact ors such as state

funding programs are also im portant the federal subsidies

granted to me dical schools beginning in the mid-sixties for both

new construction and expansion and as a direct incentive to

expand enrollment (per student capitation grants ) may be viewed

as a summary mea sure of forces weakening the AMAs control over

US me dical schools That is if the AMA as discussed earlier

provided one force which inhibited medical schools from expandshy

ing then the federal government may have supplied an opposing

influence

Actual entry of the dominant firm (DFNL ) in any given year t

is posited to equal a proportion y of the di fference between

the anticipated optimal stock in year t MDSt and the sum of

year t-ls anticipated actual stock MDSt-lbull and the anticipated

fringe entry in year t CNLt pl us a suf ficient number to

replace fully the anticipated depreciation of last years

stock10 That is

Previ ously each organization undertook se parate school surveys

lO The rate of depreciation is assumed to be constant across time and known

9

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

population) in 1981 after remaining fairly stable since the end

o f World War II The number of Foreign Medical Graduates (FMG)

ac hieving initial US licensure annually more than quadr upled

between 1965 and 19722 Similarly US medical school output

doubled between 1965 and 1980 Can all of this growth in the

number of physicians practicing in the us be explained by an

expansion of demand due to increased government expenditures for

medical care (eg Medicare and Medicaid) as well as hig her

consumer incomes Or is it also tr ue that the AMA has become

less effective at maintaining the entry barriers necessary to

restrict the supply of physicians permitted to practice in the

United States Evidence exists to suggest that the ret urn to

p hysician training has decreased substantially over the la st

decade Figure 1 portrays the ratio of physician income to

opportunity cost3 over the last 37 years While some of that

decline is undoubtedly attributable to a grad ual return to

equilibrium following a rapid and la rge increase in demand in the

1960s the downward trend has continued long enoug h to arouse

suspicions of permanent changes in market structure Moreover

the popular and medical trade press provides numerous indications

2 FMG entry has fallen in the last eig ht years but is still double its 1965 level

3 Opportunity cost is measured wit h college graduate earnings adjusted for additional training time as the benc hmark Its derivation is described in detail in Noether supra note 1

- 2shy

l

f I

-middot ---------------0----

145

1 bullbull o

I J5

1 s

llO

I _ _

_ _ ___ ___ _ _ __ __ _ _ _ _ ampamp n C 12 - a ao

F I GURE l

Phy s i c i a n I ncome - Oppo r tu n i ty Cos t Rat i o 1 9 4 6- 1 9 8 2

lNCrOC 50

I I

I

lO

I I I

bull 15

110

bullamp bulla so sz sbull bull

3 --

--------------------- - ----

that physicians themselves feel that they are operating in a more

competitive envir onment4

AMA authority over the medical sc hool accreditation process

combined wit h mandatory state licensure of individual physicians

has been cited as an effective restriction of the physician

supplyS The AMA has not only controlled the supply of US

medical sc hool spaces by making it di fficult for new sc hools to

enter and by preventing expansion of existing sc hools6 b ut has

also worked to assure that state licensing statutes require

graduation from AMA-accredited sc hools Foreign entry has al so

been curtailed by restrictive licensing laws as well as a strict

federal immigration policy However some of these policies have

been modified in recent years Have such changes led to

increased entry into the medical profession

Despite AMA pr otestations that the supply of US grad uates

was adequate public concern about a physician shortage

developed during the 1 950s A fter several failures in the mid shy

1960s bot h federal and state governments ena cted funding

programs to encourage expansion of existing and creation of ne w

4 See for example Coping with the Approaching Doctor Glut New England Journal of Medicine (1982) cited in AMA to Help MDs Adjust to Market American Medical News 12 November 1982 at 1 and Doctors are Entering a Brave New World of Competition B usiness Week 16 July 1984 at 56-61

5 See for example Kessel supra note 1 Friedman supra note 1 and D R Hyde P Wol ff A Gross and E L Hof fman The American Medical Association Power Purpose and Politics in Organized Medicine 6 3 Yale Law Journal 938 (1954)

6 Schools are accredited for a given size level

- 4shy

medical schools 7 At ab out the same time the federal immigrashy

tion law was eased and thr oughout the 1960s state licensing laws

were relaxed It appears that policies designed to inhibit

growth in the supply of physicians practicing in the United

States have been weakened If these programs were effective

then we should be able to explain at least part of the recent

expansion of the physician stock by the re gulatory and funding

c hanges that have occurred Such evidence would suggest that AMA

authority has dimini shed 8

7 The AMAs position on public funding of medical scho ols changed during the 1 950s from one of total opposition to acknowshyledgment that a one-time grant to support construction prefershyably donated by the private sector would be beneficial During this decade both medical sc hools whic h ca me to rely increasingly on federal research funding and hospit als whose expansion increased demand for medical residents favored strong federal support of medical schools This pressure coupled with that from the general public for increased physician supp ly as well as rising physician incomes during the 1 950s whic h signalled increased demand for medical ser vices perhaps led to the AMAs approval by 1960 o f a controlled expansion of medical schools The funding programs subsequently enacted at both federal and state le vels however far exceeded the AMAs desires For a detailed acco unt of the AMAs ef f orts to thwart the enactment of federal programs to aid medical schools see Elton Rayack Professional Power and American Medicine The Economics of the American Medical Association (1967)

8 Tho mas Hall and Cotton Lindsay Medical Schools Producers of W hat Sellers to W hom 23 Journal of Law and Economics 55 ( 1980) at 58 and 78 cite the gr owth in us licensing of FMGs as

well as medical sc hools positive enrollment response to donashytions particularly federal as evidence that the medical professhys ion possesses no monopoly power M ost notably the pr ofess ion [organized me dicine] has been spectacuarly unsuccessful in preventing expansion of su pply through immigration of foreignshytrained physician s Have me dical schools co nspired with organized medicine to restrict the supply of practitioners Our evidence suggests that they have not Medical sc hool output is

(f ootnote co ntinued)

-5shy

In this paper we model the stock of US practicing physishy

cians as composed of a dominant firm of AMA-sanctioned physicians

and a competitive fringe whose entry is imper fectly restricted

C hanges in the degree of competition in the physician market are

measured by estimating the gr owth of the competitive fringe

component due to the relaxation of vari ous restrictions A s

noted above until the advent of large -scale federal involvement

in medicine in the 1960s it is possible to view us physicians

as a dominant firm that maintained entry barriers pr om ulgated by

the AMA and enforced by federal and state governments su ffishy

ciently hig h to prevent most competitive fringe physicians from

entering The tremendous growth in the physician -population

ratio in the la st 15 yea rs however can per haps be explained by

the increased entry of the competitive fringe due to eroding

entry barriers

As suggested above two sources of a growing competitive

fringe seem likely First de spite AMA ef for ts substantial

numbers of foreign medical gr aduates have been licensed in the

United States part icularly in the last 15 years since many

restrictive laws have been relaxed They certainly pr ovide a

product similar en ough to be considered a substitute Second

the number of us me dical students gr ad uating ann ually has risen

(footnote continues)

positively related to demand by both donors and applicants Interestingly they never seem to con sider the possibility that the AMAs power may have weakened over time

-6shy

-7-

sharply While this may be explained solely by a growth in

demand it is also possible that the AMAs control over the

number of medical school spaces weakened lea ding to the cr eation

o f a domestic competitive fringe The Liaison Committee on

Hedical Education established in 1942 by the AMA and the

American Association of Medical Colleges9 now contains represen shy

tatives for the public and federal government in addition to its

founding organizations While other fact ors such as state

funding programs are also im portant the federal subsidies

granted to me dical schools beginning in the mid-sixties for both

new construction and expansion and as a direct incentive to

expand enrollment (per student capitation grants ) may be viewed

as a summary mea sure of forces weakening the AMAs control over

US me dical schools That is if the AMA as discussed earlier

provided one force which inhibited medical schools from expandshy

ing then the federal government may have supplied an opposing

influence

Actual entry of the dominant firm (DFNL ) in any given year t

is posited to equal a proportion y of the di fference between

the anticipated optimal stock in year t MDSt and the sum of

year t-ls anticipated actual stock MDSt-lbull and the anticipated

fringe entry in year t CNLt pl us a suf ficient number to

replace fully the anticipated depreciation of last years

stock10 That is

Previ ously each organization undertook se parate school surveys

lO The rate of depreciation is assumed to be constant across time and known

9

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

l

f I

-middot ---------------0----

145

1 bullbull o

I J5

1 s

llO

I _ _

_ _ ___ ___ _ _ __ __ _ _ _ _ ampamp n C 12 - a ao

F I GURE l

Phy s i c i a n I ncome - Oppo r tu n i ty Cos t Rat i o 1 9 4 6- 1 9 8 2

lNCrOC 50

I I

I

lO

I I I

bull 15

110

bullamp bulla so sz sbull bull

3 --

--------------------- - ----

that physicians themselves feel that they are operating in a more

competitive envir onment4

AMA authority over the medical sc hool accreditation process

combined wit h mandatory state licensure of individual physicians

has been cited as an effective restriction of the physician

supplyS The AMA has not only controlled the supply of US

medical sc hool spaces by making it di fficult for new sc hools to

enter and by preventing expansion of existing sc hools6 b ut has

also worked to assure that state licensing statutes require

graduation from AMA-accredited sc hools Foreign entry has al so

been curtailed by restrictive licensing laws as well as a strict

federal immigration policy However some of these policies have

been modified in recent years Have such changes led to

increased entry into the medical profession

Despite AMA pr otestations that the supply of US grad uates

was adequate public concern about a physician shortage

developed during the 1 950s A fter several failures in the mid shy

1960s bot h federal and state governments ena cted funding

programs to encourage expansion of existing and creation of ne w

4 See for example Coping with the Approaching Doctor Glut New England Journal of Medicine (1982) cited in AMA to Help MDs Adjust to Market American Medical News 12 November 1982 at 1 and Doctors are Entering a Brave New World of Competition B usiness Week 16 July 1984 at 56-61

5 See for example Kessel supra note 1 Friedman supra note 1 and D R Hyde P Wol ff A Gross and E L Hof fman The American Medical Association Power Purpose and Politics in Organized Medicine 6 3 Yale Law Journal 938 (1954)

6 Schools are accredited for a given size level

- 4shy

medical schools 7 At ab out the same time the federal immigrashy

tion law was eased and thr oughout the 1960s state licensing laws

were relaxed It appears that policies designed to inhibit

growth in the supply of physicians practicing in the United

States have been weakened If these programs were effective

then we should be able to explain at least part of the recent

expansion of the physician stock by the re gulatory and funding

c hanges that have occurred Such evidence would suggest that AMA

authority has dimini shed 8

7 The AMAs position on public funding of medical scho ols changed during the 1 950s from one of total opposition to acknowshyledgment that a one-time grant to support construction prefershyably donated by the private sector would be beneficial During this decade both medical sc hools whic h ca me to rely increasingly on federal research funding and hospit als whose expansion increased demand for medical residents favored strong federal support of medical schools This pressure coupled with that from the general public for increased physician supp ly as well as rising physician incomes during the 1 950s whic h signalled increased demand for medical ser vices perhaps led to the AMAs approval by 1960 o f a controlled expansion of medical schools The funding programs subsequently enacted at both federal and state le vels however far exceeded the AMAs desires For a detailed acco unt of the AMAs ef f orts to thwart the enactment of federal programs to aid medical schools see Elton Rayack Professional Power and American Medicine The Economics of the American Medical Association (1967)

8 Tho mas Hall and Cotton Lindsay Medical Schools Producers of W hat Sellers to W hom 23 Journal of Law and Economics 55 ( 1980) at 58 and 78 cite the gr owth in us licensing of FMGs as

well as medical sc hools positive enrollment response to donashytions particularly federal as evidence that the medical professhys ion possesses no monopoly power M ost notably the pr ofess ion [organized me dicine] has been spectacuarly unsuccessful in preventing expansion of su pply through immigration of foreignshytrained physician s Have me dical schools co nspired with organized medicine to restrict the supply of practitioners Our evidence suggests that they have not Medical sc hool output is

(f ootnote co ntinued)

-5shy

In this paper we model the stock of US practicing physishy

cians as composed of a dominant firm of AMA-sanctioned physicians

and a competitive fringe whose entry is imper fectly restricted

C hanges in the degree of competition in the physician market are

measured by estimating the gr owth of the competitive fringe

component due to the relaxation of vari ous restrictions A s

noted above until the advent of large -scale federal involvement

in medicine in the 1960s it is possible to view us physicians

as a dominant firm that maintained entry barriers pr om ulgated by

the AMA and enforced by federal and state governments su ffishy

ciently hig h to prevent most competitive fringe physicians from

entering The tremendous growth in the physician -population

ratio in the la st 15 yea rs however can per haps be explained by

the increased entry of the competitive fringe due to eroding

entry barriers

As suggested above two sources of a growing competitive

fringe seem likely First de spite AMA ef for ts substantial

numbers of foreign medical gr aduates have been licensed in the

United States part icularly in the last 15 years since many

restrictive laws have been relaxed They certainly pr ovide a

product similar en ough to be considered a substitute Second

the number of us me dical students gr ad uating ann ually has risen

(footnote continues)

positively related to demand by both donors and applicants Interestingly they never seem to con sider the possibility that the AMAs power may have weakened over time

-6shy

-7-

sharply While this may be explained solely by a growth in

demand it is also possible that the AMAs control over the

number of medical school spaces weakened lea ding to the cr eation

o f a domestic competitive fringe The Liaison Committee on

Hedical Education established in 1942 by the AMA and the

American Association of Medical Colleges9 now contains represen shy

tatives for the public and federal government in addition to its

founding organizations While other fact ors such as state

funding programs are also im portant the federal subsidies

granted to me dical schools beginning in the mid-sixties for both

new construction and expansion and as a direct incentive to

expand enrollment (per student capitation grants ) may be viewed

as a summary mea sure of forces weakening the AMAs control over

US me dical schools That is if the AMA as discussed earlier

provided one force which inhibited medical schools from expandshy

ing then the federal government may have supplied an opposing

influence

Actual entry of the dominant firm (DFNL ) in any given year t

is posited to equal a proportion y of the di fference between

the anticipated optimal stock in year t MDSt and the sum of

year t-ls anticipated actual stock MDSt-lbull and the anticipated

fringe entry in year t CNLt pl us a suf ficient number to

replace fully the anticipated depreciation of last years

stock10 That is

Previ ously each organization undertook se parate school surveys

lO The rate of depreciation is assumed to be constant across time and known

9

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

--------------------- - ----

that physicians themselves feel that they are operating in a more

competitive envir onment4

AMA authority over the medical sc hool accreditation process

combined wit h mandatory state licensure of individual physicians

has been cited as an effective restriction of the physician

supplyS The AMA has not only controlled the supply of US

medical sc hool spaces by making it di fficult for new sc hools to

enter and by preventing expansion of existing sc hools6 b ut has

also worked to assure that state licensing statutes require

graduation from AMA-accredited sc hools Foreign entry has al so

been curtailed by restrictive licensing laws as well as a strict

federal immigration policy However some of these policies have

been modified in recent years Have such changes led to

increased entry into the medical profession

Despite AMA pr otestations that the supply of US grad uates

was adequate public concern about a physician shortage

developed during the 1 950s A fter several failures in the mid shy

1960s bot h federal and state governments ena cted funding

programs to encourage expansion of existing and creation of ne w

4 See for example Coping with the Approaching Doctor Glut New England Journal of Medicine (1982) cited in AMA to Help MDs Adjust to Market American Medical News 12 November 1982 at 1 and Doctors are Entering a Brave New World of Competition B usiness Week 16 July 1984 at 56-61

5 See for example Kessel supra note 1 Friedman supra note 1 and D R Hyde P Wol ff A Gross and E L Hof fman The American Medical Association Power Purpose and Politics in Organized Medicine 6 3 Yale Law Journal 938 (1954)

6 Schools are accredited for a given size level

- 4shy

medical schools 7 At ab out the same time the federal immigrashy

tion law was eased and thr oughout the 1960s state licensing laws

were relaxed It appears that policies designed to inhibit

growth in the supply of physicians practicing in the United

States have been weakened If these programs were effective

then we should be able to explain at least part of the recent

expansion of the physician stock by the re gulatory and funding

c hanges that have occurred Such evidence would suggest that AMA

authority has dimini shed 8

7 The AMAs position on public funding of medical scho ols changed during the 1 950s from one of total opposition to acknowshyledgment that a one-time grant to support construction prefershyably donated by the private sector would be beneficial During this decade both medical sc hools whic h ca me to rely increasingly on federal research funding and hospit als whose expansion increased demand for medical residents favored strong federal support of medical schools This pressure coupled with that from the general public for increased physician supp ly as well as rising physician incomes during the 1 950s whic h signalled increased demand for medical ser vices perhaps led to the AMAs approval by 1960 o f a controlled expansion of medical schools The funding programs subsequently enacted at both federal and state le vels however far exceeded the AMAs desires For a detailed acco unt of the AMAs ef f orts to thwart the enactment of federal programs to aid medical schools see Elton Rayack Professional Power and American Medicine The Economics of the American Medical Association (1967)

8 Tho mas Hall and Cotton Lindsay Medical Schools Producers of W hat Sellers to W hom 23 Journal of Law and Economics 55 ( 1980) at 58 and 78 cite the gr owth in us licensing of FMGs as

well as medical sc hools positive enrollment response to donashytions particularly federal as evidence that the medical professhys ion possesses no monopoly power M ost notably the pr ofess ion [organized me dicine] has been spectacuarly unsuccessful in preventing expansion of su pply through immigration of foreignshytrained physician s Have me dical schools co nspired with organized medicine to restrict the supply of practitioners Our evidence suggests that they have not Medical sc hool output is

(f ootnote co ntinued)

-5shy

In this paper we model the stock of US practicing physishy

cians as composed of a dominant firm of AMA-sanctioned physicians

and a competitive fringe whose entry is imper fectly restricted

C hanges in the degree of competition in the physician market are

measured by estimating the gr owth of the competitive fringe

component due to the relaxation of vari ous restrictions A s

noted above until the advent of large -scale federal involvement

in medicine in the 1960s it is possible to view us physicians

as a dominant firm that maintained entry barriers pr om ulgated by

the AMA and enforced by federal and state governments su ffishy

ciently hig h to prevent most competitive fringe physicians from

entering The tremendous growth in the physician -population

ratio in the la st 15 yea rs however can per haps be explained by

the increased entry of the competitive fringe due to eroding

entry barriers

As suggested above two sources of a growing competitive

fringe seem likely First de spite AMA ef for ts substantial

numbers of foreign medical gr aduates have been licensed in the

United States part icularly in the last 15 years since many

restrictive laws have been relaxed They certainly pr ovide a

product similar en ough to be considered a substitute Second

the number of us me dical students gr ad uating ann ually has risen

(footnote continues)

positively related to demand by both donors and applicants Interestingly they never seem to con sider the possibility that the AMAs power may have weakened over time

-6shy

-7-

sharply While this may be explained solely by a growth in

demand it is also possible that the AMAs control over the

number of medical school spaces weakened lea ding to the cr eation

o f a domestic competitive fringe The Liaison Committee on

Hedical Education established in 1942 by the AMA and the

American Association of Medical Colleges9 now contains represen shy

tatives for the public and federal government in addition to its

founding organizations While other fact ors such as state

funding programs are also im portant the federal subsidies

granted to me dical schools beginning in the mid-sixties for both

new construction and expansion and as a direct incentive to

expand enrollment (per student capitation grants ) may be viewed

as a summary mea sure of forces weakening the AMAs control over

US me dical schools That is if the AMA as discussed earlier

provided one force which inhibited medical schools from expandshy

ing then the federal government may have supplied an opposing

influence

Actual entry of the dominant firm (DFNL ) in any given year t

is posited to equal a proportion y of the di fference between

the anticipated optimal stock in year t MDSt and the sum of

year t-ls anticipated actual stock MDSt-lbull and the anticipated

fringe entry in year t CNLt pl us a suf ficient number to

replace fully the anticipated depreciation of last years

stock10 That is

Previ ously each organization undertook se parate school surveys

lO The rate of depreciation is assumed to be constant across time and known

9

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

medical schools 7 At ab out the same time the federal immigrashy

tion law was eased and thr oughout the 1960s state licensing laws

were relaxed It appears that policies designed to inhibit

growth in the supply of physicians practicing in the United

States have been weakened If these programs were effective

then we should be able to explain at least part of the recent

expansion of the physician stock by the re gulatory and funding

c hanges that have occurred Such evidence would suggest that AMA

authority has dimini shed 8

7 The AMAs position on public funding of medical scho ols changed during the 1 950s from one of total opposition to acknowshyledgment that a one-time grant to support construction prefershyably donated by the private sector would be beneficial During this decade both medical sc hools whic h ca me to rely increasingly on federal research funding and hospit als whose expansion increased demand for medical residents favored strong federal support of medical schools This pressure coupled with that from the general public for increased physician supp ly as well as rising physician incomes during the 1 950s whic h signalled increased demand for medical ser vices perhaps led to the AMAs approval by 1960 o f a controlled expansion of medical schools The funding programs subsequently enacted at both federal and state le vels however far exceeded the AMAs desires For a detailed acco unt of the AMAs ef f orts to thwart the enactment of federal programs to aid medical schools see Elton Rayack Professional Power and American Medicine The Economics of the American Medical Association (1967)

8 Tho mas Hall and Cotton Lindsay Medical Schools Producers of W hat Sellers to W hom 23 Journal of Law and Economics 55 ( 1980) at 58 and 78 cite the gr owth in us licensing of FMGs as

well as medical sc hools positive enrollment response to donashytions particularly federal as evidence that the medical professhys ion possesses no monopoly power M ost notably the pr ofess ion [organized me dicine] has been spectacuarly unsuccessful in preventing expansion of su pply through immigration of foreignshytrained physician s Have me dical schools co nspired with organized medicine to restrict the supply of practitioners Our evidence suggests that they have not Medical sc hool output is

(f ootnote co ntinued)

-5shy

In this paper we model the stock of US practicing physishy

cians as composed of a dominant firm of AMA-sanctioned physicians

and a competitive fringe whose entry is imper fectly restricted

C hanges in the degree of competition in the physician market are

measured by estimating the gr owth of the competitive fringe

component due to the relaxation of vari ous restrictions A s

noted above until the advent of large -scale federal involvement

in medicine in the 1960s it is possible to view us physicians

as a dominant firm that maintained entry barriers pr om ulgated by

the AMA and enforced by federal and state governments su ffishy

ciently hig h to prevent most competitive fringe physicians from

entering The tremendous growth in the physician -population

ratio in the la st 15 yea rs however can per haps be explained by

the increased entry of the competitive fringe due to eroding

entry barriers

As suggested above two sources of a growing competitive

fringe seem likely First de spite AMA ef for ts substantial

numbers of foreign medical gr aduates have been licensed in the

United States part icularly in the last 15 years since many

restrictive laws have been relaxed They certainly pr ovide a

product similar en ough to be considered a substitute Second

the number of us me dical students gr ad uating ann ually has risen

(footnote continues)

positively related to demand by both donors and applicants Interestingly they never seem to con sider the possibility that the AMAs power may have weakened over time

-6shy

-7-

sharply While this may be explained solely by a growth in

demand it is also possible that the AMAs control over the

number of medical school spaces weakened lea ding to the cr eation

o f a domestic competitive fringe The Liaison Committee on

Hedical Education established in 1942 by the AMA and the

American Association of Medical Colleges9 now contains represen shy

tatives for the public and federal government in addition to its

founding organizations While other fact ors such as state

funding programs are also im portant the federal subsidies

granted to me dical schools beginning in the mid-sixties for both

new construction and expansion and as a direct incentive to

expand enrollment (per student capitation grants ) may be viewed

as a summary mea sure of forces weakening the AMAs control over

US me dical schools That is if the AMA as discussed earlier

provided one force which inhibited medical schools from expandshy

ing then the federal government may have supplied an opposing

influence

Actual entry of the dominant firm (DFNL ) in any given year t

is posited to equal a proportion y of the di fference between

the anticipated optimal stock in year t MDSt and the sum of

year t-ls anticipated actual stock MDSt-lbull and the anticipated

fringe entry in year t CNLt pl us a suf ficient number to

replace fully the anticipated depreciation of last years

stock10 That is

Previ ously each organization undertook se parate school surveys

lO The rate of depreciation is assumed to be constant across time and known

9

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

In this paper we model the stock of US practicing physishy

cians as composed of a dominant firm of AMA-sanctioned physicians

and a competitive fringe whose entry is imper fectly restricted

C hanges in the degree of competition in the physician market are

measured by estimating the gr owth of the competitive fringe

component due to the relaxation of vari ous restrictions A s

noted above until the advent of large -scale federal involvement

in medicine in the 1960s it is possible to view us physicians

as a dominant firm that maintained entry barriers pr om ulgated by

the AMA and enforced by federal and state governments su ffishy

ciently hig h to prevent most competitive fringe physicians from

entering The tremendous growth in the physician -population

ratio in the la st 15 yea rs however can per haps be explained by

the increased entry of the competitive fringe due to eroding

entry barriers

As suggested above two sources of a growing competitive

fringe seem likely First de spite AMA ef for ts substantial

numbers of foreign medical gr aduates have been licensed in the

United States part icularly in the last 15 years since many

restrictive laws have been relaxed They certainly pr ovide a

product similar en ough to be considered a substitute Second

the number of us me dical students gr ad uating ann ually has risen

(footnote continues)

positively related to demand by both donors and applicants Interestingly they never seem to con sider the possibility that the AMAs power may have weakened over time

-6shy

-7-

sharply While this may be explained solely by a growth in

demand it is also possible that the AMAs control over the

number of medical school spaces weakened lea ding to the cr eation

o f a domestic competitive fringe The Liaison Committee on

Hedical Education established in 1942 by the AMA and the

American Association of Medical Colleges9 now contains represen shy

tatives for the public and federal government in addition to its

founding organizations While other fact ors such as state

funding programs are also im portant the federal subsidies

granted to me dical schools beginning in the mid-sixties for both

new construction and expansion and as a direct incentive to

expand enrollment (per student capitation grants ) may be viewed

as a summary mea sure of forces weakening the AMAs control over

US me dical schools That is if the AMA as discussed earlier

provided one force which inhibited medical schools from expandshy

ing then the federal government may have supplied an opposing

influence

Actual entry of the dominant firm (DFNL ) in any given year t

is posited to equal a proportion y of the di fference between

the anticipated optimal stock in year t MDSt and the sum of

year t-ls anticipated actual stock MDSt-lbull and the anticipated

fringe entry in year t CNLt pl us a suf ficient number to

replace fully the anticipated depreciation of last years

stock10 That is

Previ ously each organization undertook se parate school surveys

lO The rate of depreciation is assumed to be constant across time and known

9

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

-7-

sharply While this may be explained solely by a growth in

demand it is also possible that the AMAs control over the

number of medical school spaces weakened lea ding to the cr eation

o f a domestic competitive fringe The Liaison Committee on

Hedical Education established in 1942 by the AMA and the

American Association of Medical Colleges9 now contains represen shy

tatives for the public and federal government in addition to its

founding organizations While other fact ors such as state

funding programs are also im portant the federal subsidies

granted to me dical schools beginning in the mid-sixties for both

new construction and expansion and as a direct incentive to

expand enrollment (per student capitation grants ) may be viewed

as a summary mea sure of forces weakening the AMAs control over

US me dical schools That is if the AMA as discussed earlier

provided one force which inhibited medical schools from expandshy

ing then the federal government may have supplied an opposing

influence

Actual entry of the dominant firm (DFNL ) in any given year t

is posited to equal a proportion y of the di fference between

the anticipated optimal stock in year t MDSt and the sum of

year t-ls anticipated actual stock MDSt-lbull and the anticipated

fringe entry in year t CNLt pl us a suf ficient number to

replace fully the anticipated depreciation of last years

stock10 That is

Previ ously each organization undertook se parate school surveys

lO The rate of depreciation is assumed to be constant across time and known

9

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

(1) DFNLt = y (MDS - (MDS + CNL ) + d MDS tlt-5 t-llt-5 tlt-5 t-llt-5

Everything is expressed in 5-year expectation horizons due to the

time required between entry into medical sc hool and actual

licensure MDS is the optimal total stock determined by the

dominant firms equation of its resid ual marginal revenue and

marginal cost curves It depends on exogenous demand and cost

parameters as well as ent ry barrier variables Fringe entry is

assumed to exhibit a supply curve that is upward sloping with

respect to US physician income and whose level is affected by

both entry barriers (R) and the fringes own oppor tunity cost

(FTOC) whic h together determine fringe marginal cost as shown

in ( 2)

I (2) CNLt = bQ + bl MDINCt + b2 R t + b3FTOCt

Figure 2 portrays the market equilibrium given the

simplifying assumptions of a linear market demand curve and

horizontal dominant firm marginal cost curvell

Section II analyzes the determinants of incr eased FMG entry

w hile Section III disc usses the effect of us medical sc hool

sub sidies on the development of a domestic fringe Section IV

provides a summary

11 A sim ple al gebraic model using these assumptions can be found in Appendix A For a complete derivation of reduced form stock income and fringe supply equations see Monica Noether The Gr owing Supply of Physicians Are the Entry Barriers Breaking (unpublished PhD Dissertation University of Chicago 1983)

or write to the aut hor

-8shy

F I G U RE 2

E qui librium Phy sician Stock and I ncome U nde r Domi na n t Firm - Compe titive Fringe Mode l

HDINC ( $I1D)

1DEgt1C T

Residual Demand(DF)

Harket Demand

HDST

DF Domina nt Firm C F C ompetitive Fri n g e

T Tot a l

9 --

i2 RF t

I I FORE I GN MED ICAL GRADUATE FR I NGE ENTRY

F i gu re 3 shows the annu a l propor t i on of tot a l new us

l i ce nsees th a t FMG s have rep re se n ted for the las t 30 yea rs

A f ter r i s i ng s low ly unt i l the m i d - 1 9 6 0 s th i s propor t i on

i nc re a sed rap i d l y i n the late - 1 9 6 0 s and ea r ly 1 9 7 0 s re a c h i n g a

peak of a lmo s t 5 0 pe rce nt in 1 9 7 2 I t the n be gan to d im i n i s h

and in the las t f i ve yea rs has fa l le n sharp l y F rom ( 2 ) we ca n

i s o la te the FMG fr i nge supp ly curve as

( 3 ) F MGt = f o + f 1 MD I NC t + + f) FMGTOC t bull

FMG supp l y is d i re c t ly re la ted to us phy s i c i a n income (MD I NC )

I t is inve rse l y as s oc i ated w i t h a se t of re s t r i c t i ve ent ry

b a r r iers a ime d at fore i gn -tra i ne d phy s i c i ans (RF ) and FMG

oppor t u n i t y cos t (FMGTOC ) wh i ch re duce the qua nt i t y supp l i e d at

a ny gi ve n income Rea l us phy s i c i a n i n come s increased by

a lmos t 5 0 pe rce nt f rom 19 6 3 to 1 9 7 2 due mos t proba b ly to the

l a rge not comp lete l y ant i c i pated growth in demand oc c a s i oned by

the enac tme n t of fede r a l insu r a n ce prog r am s Thu s growth in FMG

e n t ry in the late 1 9 6 0 s and ea r l y 1 9 7 0 s cou l d have re su l te d

e nt i re ly from h i ghe r us phy s i c i a n income s and have bee n due

s o le ly to moveme nt along a s i ng le supp ly cu rve

A l te rna t i ve ly the i n c re a se i n FMG ent ry may a l s o be

a t t r i buted to the re l a xa t i on of ma ny ent ry ba r r i e rs du r i n g th i s

per i od wh i ch sh i f te d the supp ly cu rve to the r i g h t I n 1 9 68 the

nat i o n a l or i g i n quo t a sy s tem of the Imm i g r a t i o n Ac t wh i ch favored

Wes te r n Hem i sp he re cou nt r i es was abo l i s he d a l low i ng ma ny Eas te rn

- 1 0 shy

I

IIIIlCH I I

Cl bull

F i g u re 3

FMG - New us L i ce nsees Ra t i o

o 7

o bull

I i

03

obull

I I

c

01

I

i I

oc

bull

middot bull4 sc 52 sbull 5amp se 60 62 i4 sa ampa 10 12 7bull 1 1e 30 YpoundoA

- 1 1 shy

Hemisphere physicians to emigrate Additionally preference

categories for those with exceptional ability or with skills in

s hort supply in the US were created and given a special extra

quota of 17000 annually Congress apparently reco gnized the

US p hysician shortage and allowed doctors to quali fy for

both of these categories From 1965 to 197 1 the numb er of FMG

immigrants more than tripled and the pr opor tion emigrating from

A sian nations rose from 13 to 73 percent 12 In 1970 the

E xc hange Visitor program established in 1948 was al so broadened

to allow anyone possessing an exchange or J visa for

e xample foreign physici ans training in US residency pr ograms

to apply for immediate change to immigrant statusl3 This

expansion took effect on July 1 197 1

Moreover many state medical licensure requirements were

relaxed during this time In 1950 15 states did not license

FMGs under any circ umstances an other 18 required us citizenshy

s hip By 1960 only si x states did not license FMGs but 21

required citizens hip In 1970 all stat es licensed FMGs an d only

eig ht required citizenship In 1973 the Supreme Court ruled

Alfonso Mejia Helena Piz urki and Erica Royston Foreign Medical Graduates (1980) at 187

l 3 Formerly a tw o-year absence from the US was req uired As Mejia et al supra note 12 notes the AMA opposed this policy c hange- shy

- 12shy

12

i2 RF t

I I FORE I GN MED ICAL GRADUATE FR I NGE ENTRY

F i gu re 3 shows the annu a l propor t i on of tot a l new us

l i ce nsees th a t FMG s have rep re se n ted for the las t 30 yea rs

A f ter r i s i ng s low ly unt i l the m i d - 1 9 6 0 s th i s propor t i on

i nc re a sed rap i d l y i n the late - 1 9 6 0 s and ea r ly 1 9 7 0 s re a c h i n g a

peak of a lmo s t 5 0 pe rce nt in 1 9 7 2 I t the n be gan to d im i n i s h

and in the las t f i ve yea rs has fa l le n sharp l y F rom ( 2 ) we ca n

i s o la te the FMG fr i nge supp ly curve as

( 3 ) F MGt = f o + f 1 MD I NC t + + f) FMGTOC t bull

FMG supp l y is d i re c t ly re la ted to us phy s i c i a n income (MD I NC )

I t is inve rse l y as s oc i ated w i t h a se t of re s t r i c t i ve ent ry

b a r r iers a ime d at fore i gn -tra i ne d phy s i c i ans (RF ) and FMG

oppor t u n i t y cos t (FMGTOC ) wh i ch re duce the qua nt i t y supp l i e d at

a ny gi ve n income Rea l us phy s i c i a n i n come s increased by

a lmos t 5 0 pe rce nt f rom 19 6 3 to 1 9 7 2 due mos t proba b ly to the

l a rge not comp lete l y ant i c i pated growth in demand oc c a s i oned by

the enac tme n t of fede r a l insu r a n ce prog r am s Thu s growth in FMG

e n t ry in the late 1 9 6 0 s and ea r l y 1 9 7 0 s cou l d have re su l te d

e nt i re ly from h i ghe r us phy s i c i a n income s and have bee n due

s o le ly to moveme nt along a s i ng le supp ly cu rve

A l te rna t i ve ly the i n c re a se i n FMG ent ry may a l s o be

a t t r i buted to the re l a xa t i on of ma ny ent ry ba r r i e rs du r i n g th i s

per i od wh i ch sh i f te d the supp ly cu rve to the r i g h t I n 1 9 68 the

nat i o n a l or i g i n quo t a sy s tem of the Imm i g r a t i o n Ac t wh i ch favored

Wes te r n Hem i sp he re cou nt r i es was abo l i s he d a l low i ng ma ny Eas te rn

- 1 0 shy

I

IIIIlCH I I

Cl bull

F i g u re 3

FMG - New us L i ce nsees Ra t i o

o 7

o bull

I i

03

obull

I I

c

01

I

i I

oc

bull

middot bull4 sc 52 sbull 5amp se 60 62 i4 sa ampa 10 12 7bull 1 1e 30 YpoundoA

- 1 1 shy

Hemisphere physicians to emigrate Additionally preference

categories for those with exceptional ability or with skills in

s hort supply in the US were created and given a special extra

quota of 17000 annually Congress apparently reco gnized the

US p hysician shortage and allowed doctors to quali fy for

both of these categories From 1965 to 197 1 the numb er of FMG

immigrants more than tripled and the pr opor tion emigrating from

A sian nations rose from 13 to 73 percent 12 In 1970 the

E xc hange Visitor program established in 1948 was al so broadened

to allow anyone possessing an exchange or J visa for

e xample foreign physici ans training in US residency pr ograms

to apply for immediate change to immigrant statusl3 This

expansion took effect on July 1 197 1

Moreover many state medical licensure requirements were

relaxed during this time In 1950 15 states did not license

FMGs under any circ umstances an other 18 required us citizenshy

s hip By 1960 only si x states did not license FMGs but 21

required citizens hip In 1970 all stat es licensed FMGs an d only

eig ht required citizenship In 1973 the Supreme Court ruled

Alfonso Mejia Helena Piz urki and Erica Royston Foreign Medical Graduates (1980) at 187

l 3 Formerly a tw o-year absence from the US was req uired As Mejia et al supra note 12 notes the AMA opposed this policy c hange- shy

- 12shy

12

I

IIIIlCH I I

Cl bull

F i g u re 3

FMG - New us L i ce nsees Ra t i o

o 7

o bull

I i

03

obull

I I

c

01

I

i I

oc

bull

middot bull4 sc 52 sbull 5amp se 60 62 i4 sa ampa 10 12 7bull 1 1e 30 YpoundoA

- 1 1 shy

Hemisphere physicians to emigrate Additionally preference

categories for those with exceptional ability or with skills in

s hort supply in the US were created and given a special extra

quota of 17000 annually Congress apparently reco gnized the

US p hysician shortage and allowed doctors to quali fy for

both of these categories From 1965 to 197 1 the numb er of FMG

immigrants more than tripled and the pr opor tion emigrating from

A sian nations rose from 13 to 73 percent 12 In 1970 the

E xc hange Visitor program established in 1948 was al so broadened

to allow anyone possessing an exchange or J visa for

e xample foreign physici ans training in US residency pr ograms

to apply for immediate change to immigrant statusl3 This

expansion took effect on July 1 197 1

Moreover many state medical licensure requirements were

relaxed during this time In 1950 15 states did not license

FMGs under any circ umstances an other 18 required us citizenshy

s hip By 1960 only si x states did not license FMGs but 21

required citizens hip In 1970 all stat es licensed FMGs an d only

eig ht required citizenship In 1973 the Supreme Court ruled

Alfonso Mejia Helena Piz urki and Erica Royston Foreign Medical Graduates (1980) at 187

l 3 Formerly a tw o-year absence from the US was req uired As Mejia et al supra note 12 notes the AMA opposed this policy c hange- shy

- 12shy

12

Hemisphere physicians to emigrate Additionally preference

categories for those with exceptional ability or with skills in

s hort supply in the US were created and given a special extra

quota of 17000 annually Congress apparently reco gnized the

US p hysician shortage and allowed doctors to quali fy for

both of these categories From 1965 to 197 1 the numb er of FMG

immigrants more than tripled and the pr opor tion emigrating from

A sian nations rose from 13 to 73 percent 12 In 1970 the

E xc hange Visitor program established in 1948 was al so broadened

to allow anyone possessing an exchange or J visa for

e xample foreign physici ans training in US residency pr ograms

to apply for immediate change to immigrant statusl3 This

expansion took effect on July 1 197 1

Moreover many state medical licensure requirements were

relaxed during this time In 1950 15 states did not license

FMGs under any circ umstances an other 18 required us citizenshy

s hip By 1960 only si x states did not license FMGs but 21

required citizens hip In 1970 all stat es licensed FMGs an d only

eig ht required citizenship In 1973 the Supreme Court ruled

Alfonso Mejia Helena Piz urki and Erica Royston Foreign Medical Graduates (1980) at 187

l 3 Formerly a tw o-year absence from the US was req uired As Mejia et al supra note 12 notes the AMA opposed this policy c hange- shy

- 12shy

12

c i t i ze n s h i p requ i reme n t s for l i ce n s u re i l le ga l l 4 The othe r

cons t r a i nt con t a i ne d in ma ny l i cens i ng s t a tu tes con ce rns inte rn shy

s h i p and re s i de ncy tra i n i n g Wh i le mo s t states now re qu i re at

leas t one yea r eve n for us me d i ca l gra du a te s and ma ny have for

seve ra l yea rs the graduate tra i n i ng requ i reme n t s fac i ng FMG s

tend to be s t r i c te r

D im i n i s hed e n t ry of FMG s du r i ng the la s t few yea rs is

subject to the same pote n t i a l exp lana t i on s Phy s i c i a n income s

fe l l du r i ng th i s pe r i od (c a u s i n g moveme nt along a su pp ly cu rve )

Moreove r in 1 9 7 6 the Hea l t h Profe s s i ona l Edu ca t i ona l As s i s t ance

Act wa s passed (to take ef f e c t i n Janu a ry 1 9 7 7 ) wh i c h re ve rsed

s ome of the l i be r a l pol i c ie s of the ea r ly l 9 7 0 s l 5 Thu s the

s u pp l y cu rve may a l s o have sh i f ted ba ck to a more re s t r i c t i ve

1 4 Obv i ou s ly c i t i ze n s h i p requ i reme n t s do no t af fe c t us c i t i z e n s who go abroad to study and the n re t u rn Des p i te the amount of pub l i c atte n t i on the se ind i v i d u a l s cu rre n t ly re ce i ve u nt i l re ce nt ly the one s who actua l ly comp leted me d i c a l schoo l in a nothe r cou n t ry we re few In 1 9 7 2 the las t ye a r th at ci t i ze n shys h i p requ i reme n t s app l i ed i n any state 24 0 us c i t i zen FMG s p as se d state l i ce n s u re exam s 3 6 pe rce n t of a l l FMGs and 1 7 perce n t of al l new l i ce nsee accor d i ng to Ame r i ca n Med i c a l Assoc i a t i on Med i c a l L i ce n s u re Stat i s t i c s f o r 19 7 2 ( 1 9 7 3 ) a t ll

1 5 P r i or to e n t ry as imm i gr a n t s FMG s now mu s t pas s pa r t s I and I I of the Nat i ona l Boa rd of Me d i c a l Exam i ne r s Exam and show compe tence i n Engl i s h Add i t i ona l ly J-category vi s a s are gra nted for tw o yea rs on ly a f te r wh i c h t ime the ind i v i du a l mu s t re turn t o h i s or he r na t i ve cou n t ry for two yea r s be f ore app ly i n g f or pe rma ne n t re s i de nt s t a t u s in the us Moreove r vi s a s are gra nted on l y to those w i t h a wr i t ten prom i se f rom a re s i de ncy p rogram to pro v i de tra i n i ng

- 1 3 shy

s ta te S i nce state l i ce n s i n g laws have not bee n t i g h te ned it

may not have sh i f ted a l l the way ba ck to its pos i t ion i n 1 9 68 1 6

A s ou t l i ne d above two types of ent ry ba r r i e r re s t r i c t i ons

a re re le va nt to fore i gn me d i c a l gradu ates F i rs t the imm i gr a shy

t i on law was subs t ant i a l l y re l a xe d i n 1 9 7 1 after some loose n i ng

i n 1 9 68 I t was subseque n t ly t i gh tene d in 1 9 7 6 Thu s two dummy

va r i a b le s a re used to me a s u re the ef fect of imm i g r a t i on po l i c i es

The f i rs t equ a l s 1 be f o re 1 9 7 1 and 0 ot he rw i se (D7 1 ) the second

1 6 A th i rd pos s i b i l i ty cou l d be con s i de re d I f fore i gn me d i c a l gradu ate s bull opportu n i ty cos t (FMGTOC ) fe l l du r i ng the late 1 9 6 0 bull s i nc re a sed ent ry cou l d have re su l ted The oppor tu n i ty cos t of imm i grat i on to the Uni ted States is proba b ly be s t thou ght of as the i ncome an FMG cou l d have ea rned if he she had rema i ne d in h i s he r nat i ve cou n t ry p l u s any tra nspor t a t i on cos t s to the U n i ted St ates Wh i le intern a t iona l phys i c i a n i ncome dat a are not rea d i ly ava i l a b le we ca n exam i ne pe r s on a l income arou nd the wor ld assum i ng tha t it p rov i de s at lea s t some re l a t i ve i n f orma shyt i on Wh i le i n the U n i ted States re a l pe r cap i t a pe r s ona l i ncome grew by 5 3 pe r ce n t from 19 5 0 - 1 9 7 0 a compa r a b le we i gh te d shya ve ra ge income f i gu re for the wor ld grew by 1 38 pe rce nt in the s ame pe r i od whe n the we i gh t s used re f le c t the actu a l d i s t r i bu t i on o f FMG imm i g r a n t s in 1 9 68 imme d i a te l y pr i or to enactme nt of the new imm i g rat i on law Moreove r in a 1 9 68 su rve y the ra t i o of p hy s i c i a n to pe rs ona l income was hi gher in As i an count r i e s th an i n the U S equa l i ng 16 9 i n the Ph i l i pp i ne s 34 9 in Tha i la n d a nd 8 5 i n the us acco rd i ng t o Me j i a et al sup ra note 12 a t 1 3 1 The re f ore i t seems l ike l y th at the-oppor t u n i ty cos t for f ore i gn me d i ca l graduates rose du r i ng the pe r i od in que s t i on I t i s true tha t the abo l i t i on of the na t i ona l quo t a sy s tem i n the Imm i gr a t i on Act in 1 9 68 se rve d to ope n the doo r s to phy s i c i ans f rom poorer cou n t r i e s and he nce w i th lowe r oppor tu n i ty cos t s The con com i t ant exp ans i on in em i gr a t i on of As i a n FMG1s howe ve r i s the re s u l t of ea s i ng re gu l a t i on not a lowe r i ng of the i r opportu n i ty cos t

- 1 4 shy

w i l l be se t to 1 af te r 1 9 7 6 (D7 6 ) 1 7 I f re s t r i c t i ons are ef fec shy

t i ve bo th dummy va r i a b le s shou ld have neg a t i ve coe f f i c ie n t s

Three l i ce n s i ng va r i a b le s are used One mea s u res the ef fec t

o f states th a t d i d not l i ce nse FMG s unde r any c i rcum s t ance s

( NOL IC ) the se cond takes ac count o f statutes w i th c i t i ze ns h i p

requ i reme n t s (C I T I Z ) and the th i rd es t imates the impact of

gra du ate tra i n i ng t ime requ i reme n ts (TNGTM ) A l l l i ce n s i n g

v a r i ables are me a s u re d a s the we i gh ted pe rce nt age o f a l l states

w i t h the re le va nt re s t r i c t i on in ea ch yea r 18 F i gure 4 shows the

p a t te rn of the se th ree va r i a b les ove r t ime The leve l s of NOL I C

a nd C I TIZ are h i ghes t in the 1 9 5 0 s and su bsequent l y de c l i ne

the i r re l a xa t i on may exp l a i n some of the gradu a l incre a se in FMG

e nt ry du r i n g the 1 9 5 0 s and ea r ly 1 9 6 0 s TNGTM excep t for a

d i p in the 1 9 7 0 s has r i se n

The FMG -popu l a t i on ra t i o (FMGP ) is used as a de pende nt

var i ab le 1 9 Bec a u se of ava i l a b le da t a the f r i nge is mea s u re d as

17 A dummy va r i a b le equ a l to 1 be f ore 1 9 68 was su bs t i t u ted for D 7 1 in some re gre s s i on s I t was ge ne ra l l y not s i gn i f i c ant p re sumab l y e i the r be c a u se the 1 9 7 1 law cha n ge was more import ant or be cause the FMG ent ry re s p onse oc c u r re d with a lag These res u l t are not repor ted here See Noe ther sup r a note 1 1 for them

18 S i nce the m i d - 1 9 6 0 s the we i gh ts used a re the pe rce n tage of nat i ona l pe r s ona l hea l t h ca re expe nd i tu re s con t r i bu ted by ea ch s ta te P r i o r to that t ime st a te bre ak downs of hea l t h expe nd i shyt u res are not ava i l a b le so pe r s on a l i ncome i s used as a we i gh t i ng f ac to r i n s tead U s i n g i ncome as the we i ght i ng factor th rou ghout p rodu ce s s im i l a r re s u l t s

1 9 Regre s s i on s we re a l s o ru n on the leve l numbe r of FMG s as we l l a s the FMG-new l i censee s ra t i o Resu l t s d i d not di f fe r s i gn i f i c a n t l y from those reported

- 1 5 shy

bull bull bull bull

bull bull bull

bull bull

bull bull bull bull bull bull

50

bull bull bull

bull bull

FIGURE 4

FMG Re s t r i c t i on s

per ent (a) FMG licensing (+) and citizenship () requirements

I i

amp0 bull

I I I

bull bull

bull bull bull

lO

bull

o bull

10 bull bull

bull -

bull6 amp4 50 SZ bull 3i 51 iiO iZ Si ampe 70 z 7bull 7amp 79 80

YEAR

(b) FMG graduate training time requirementsyears 1 i

II

1 5 I 1

bull

bull w bull bull

bull 1 bull bull

bull o

bullCll

e

i I

01 - bull

c a I bull -cs

a f low of new ent r a n t s ( l i ce nsees ) ra the r th a n as the stock

e x i s t i ng each yea r By cumu l a t i ng the yea r l y f low ma gn i t udes

ove r the pe r i od i n que s t i on we ca n es t imate the tot a l impa c t of

the f r i nge on us phy s i c i an supp l y The FMG ent ry equ a t i on to

be es t imated i s

A

( 4 ) FMGP t = fQ + f l MD I NC t + f 2 1 D7 l t + f 2 2 D 7 6 t + f 2 3 NOL I C t

us phy s i c i a n income is not a t rue exoge nou s va r i a b le s i n ce

i t is obv i ou s ly af f e c ted by the numbe r of FMG s th at enter the

U n i ted St a te s The re f ore two -s t a ge lea s t squ a re s is used to

e s t imate the sy s tem of ( 4 ) and a reduced form income equ a t i on to

de s c r i be MD I NC th at con t a i ns va r i ables to me a s u re both fore i g n

a nd dome s t i c ent ry re s t r i c t i o n s as we l l as dema nd parame te r s and

that ac c o u nt s f or gradu a l supp ly ad j u s tme n t s to equ i l i br i um

c hange s th rou gh i n c l u s i on of the lagged stock and a pre d i c t i on

e rror te rm

2 0 I n the es t i ma t i on no va r i a b le is used to account for FMG oppor tu n i ty cos t S i nce fore i gn phy s i c i a n income da t a are not rea d i ly ava i l a b le interna t i on a l income da t a we re used to crea te a proxy f or FMGTOC A we i ghted ave ra ge of va r i ou s wor l d a re a s (eg Eas te r n As i a ) pe r cap i t a gros s na t i ona l produ c t in con s t a n t do l l a r s was ca l c u l a te d us i ng the 1 9 68 na t i ona l propor shyt i ons of FMG en t r a nts as we i gh t s A s im i la r va r i a b le was a l s o t r ied us i ng 1 9 6 5 we i g h ts B o t h showe d i na p p rop r i ate pos i t i ve e f f e c t s on FMG ent ry and proba b l y re f le cted factors not re l a ted t o FMG opp ortu n i ty cos t of imm i g r a t i n g to the us

- 1 7 shy

This re duced form income equation is de rive d from demand and

c o s t equa tions the unde r lying mode l for which is des c ribe d in

g re a te r de t ai l in Appe ndi x A I t uses pe r capita rea l in come

( INCM ) and the pe rce n t of the popu l a tion co ve re d by pu b lic insu r shy

a nce (PUB I NS ) as dem a n d va riable s 2 1 They shou l d both have a

positive e f fect on phy sician income Margina l cos t is me a s u red

as the tot a l opp ortunity cos t (TOC ) of be coming a physicia n

based on the as s ump tion tha t the ne xt be s t a l te rna tive is workin g

f o l lowin g fou r yea r s of co l le ge a n d ad j u s ted f o r the time spe nt

2 2in me di c a l schoo l and re side n cy prog r ams It shou l d a l s o be

p ositive l y co rre l a te d wit h phy sician income V aria b le s tha t

mea s u re fede ra l subsidies gra nted t o us medica l schoo l s bo th

f or con s t ru ction (CON S 8 ) and as capitation grants (CAPS 5 ) a re

a ls o in corporated and wil l be dis c u s se d fu rthe r in Section I I I

be low Whi le a true re s t ric tion on a dome s tic f ringe shou ld

r aise phy sician income since these are subsidie s they ease

e n t ry of the dome s tic f ringe and he nce lowe r phy sician in come

Fina l ly phy sician income is a f fe c te d by di f fe re n ce s be twee n

a c tu a l ent ry int o the profession and th at pre dicted by the

dominant firm As a re lated pape r shows this error ca n be

2 l O the r demand va ri a b le s we re a l s o used in Noe t he r sup r a notes 1 and 1 1 and these appe a re d mos t signific a n t Hence othe r variable s su ch as edu ca tion and age we re omit ted to avoid mu lti-co l line a rity prob lems with other independe n t va ri a b le s

2 2 The de riva tion and me a s u reme nt of TOC are fu l ly de s c ribe d in N oe t he r supra note 1 1 D at a sou rce s for TOC and the other v a ria bles used are lis ted in Appendi x B

- 1 8shy

supra

re duced to incorrect pre d i ct i ons about FMG ent ry wh i ch we de note

e F 2 3 Thu s the est i mated equ at i on i s as fol lows

( 5 ) MDINCt = oo + 0 1 TOCt + 0 2 1 INCMt + 0 2 2 PUBINSt + 0 3 1 D7 1t

+ 03 2 D7 6t + 0 3 3 NOLICt + 0 3 4 CITI Zt + 0 4 1 CON S 8t

F 2 + 04 2 CAPS 5t + 0 5 MDSPt - 1 + 0 6 et + Et

Add i t i ona l ly the re du c e d form equat i on de s cr i b i ng FMG

e nt ry de r i ved by su bst itut i ng ( 5 ) i nto ( 4 ) is a l s o est imated

a s

( 6 ) F MGPt = + TOCt + 2 1 INCMt + 2 2 PUBINSt + 3 1 D7 1t0 1

+ 3 2 D7 6t + 3 3 NOLICt + 3 4 CITI Zt + 3 5 TNGTMt

F 3 + 4 1 CON S 8t + 4 2 CAPS 5t + 5 MDSPt - 1 + 6 et + Et

The ef f e cts of the FMG re str i ct i on s and opp o rtu n i ty cost on the

redu c e d form of FMG are i ndete rm i nate The i r d i r e ct ef f e ct is

obv i ou s ly ne gat i ve but th ey have a pos it i ve i n f l u e n c e on i ncome

w h i ch a l s o a f f e cts the dep e nd e nt va r i a b l e pos it i ve l y P re s umshy

a b l y the forme r ef f e ct is strong e r so that the net impact is

negat i ve The rema i n i ng va r i a b l e s have th e same i nf luence as

they do on MDINC i e th e coe f f i c i e nts on TOC INCM PUBINS

2 3 eFSee Noethe r note 1 1 The me asu reme nt of is d e s cr i be d in Append i x B

- 1 9 shy

FCON S 8 and CAPS 5 are pos it i ve wh i l e tho s e on MDS Pt - 1 and e are

negat i ve

Column s 1 and 2 of tab l e 1 pre s e nt the re s u lts from a 2SLS

regre s s i on of the sy stem conta i n i ng the re duc e d form income

equat i on ( 5 ) and th e FMG supp ly cu rve ( 4 ) In the MDINC equat i on

a l l the fore i gn and dome st i c re str i ct i ons show the pre d i ct e d

s i gns t h e FMG ent ry re st r i ct i ons inc r e a s e phy s i c i a n income

wh i le subs i d i es to us me d i c a l schoo l s de c r e a s e it The

p re d i ct i on of phy s i c i a n i ncome de r i ved here is used as the MDINC

va r i a b l e i n the FMG equ at i on s reported here in co lumn 2 It has

the expe cted pos it i ve ef f e ct A l l of the re str i ct i on va r i a b l e s

have the ant i c i p at e d negat i ve e f f e ct on FMG ent ry 2 4

Wh i l e th e re l a xat i on o f l i ce n s i ng and imm i grat i on laws

app arent ly d i d sh i ft the FMG supp l y cu rve to the r i g ht the

coe f f i c i e nt on the D7 6 va r i a b l e a l s o su ggests th at the cu rve has

s h i fted ba c k in rec e nt ye a rs In f a ct the coe f f i c i e nt on D7 6 i s

a lmost a s large a s that on D7 1 F rom th e est i mated FMG equ at i on s

and the pre d i ct i ons o f MDINC we c a n tra ce th e moveme nt of the

F MG suppl y cu rve ove r time F i gure 5 tra c e s its moveme nt It

s hows that afte r rema i n i ng re l at i ve ly stat i o n a ry unt i l th e lat e

1 9 60 s the supp l y sh i fted out subst a nt i a l ly i n the f i rst ha l f of

the 1 9 70 s The outw a rd moveme nt cont i nu e d unt i l about 1 9 7 5 du e

2 4 E ither i n c l u d i ng TNGTM i n the i ncome equ at i on or not i nc l u d i ng it in the FMG equat i on doe s not ch ange the re su lt s s i g n i f i c a nt ly The s imp le co rre lat i on betwe e n TOC and TNGTM e xc e e d s 9

- 20 shy

s ta te S i nce state l i ce n s i n g laws have not bee n t i g h te ned it

may not have sh i f ted a l l the way ba ck to its pos i t ion i n 1 9 68 1 6

A s ou t l i ne d above two types of ent ry ba r r i e r re s t r i c t i ons

a re re le va nt to fore i gn me d i c a l gradu ates F i rs t the imm i gr a shy

t i on law was subs t ant i a l l y re l a xe d i n 1 9 7 1 after some loose n i ng

i n 1 9 68 I t was subseque n t ly t i gh tene d in 1 9 7 6 Thu s two dummy

va r i a b le s a re used to me a s u re the ef fect of imm i g r a t i on po l i c i es

The f i rs t equ a l s 1 be f o re 1 9 7 1 and 0 ot he rw i se (D7 1 ) the second

1 6 A th i rd pos s i b i l i ty cou l d be con s i de re d I f fore i gn me d i c a l gradu ate s bull opportu n i ty cos t (FMGTOC ) fe l l du r i ng the late 1 9 6 0 bull s i nc re a sed ent ry cou l d have re su l ted The oppor tu n i ty cos t of imm i grat i on to the Uni ted States is proba b ly be s t thou ght of as the i ncome an FMG cou l d have ea rned if he she had rema i ne d in h i s he r nat i ve cou n t ry p l u s any tra nspor t a t i on cos t s to the U n i ted St ates Wh i le intern a t iona l phys i c i a n i ncome dat a are not rea d i ly ava i l a b le we ca n exam i ne pe r s on a l income arou nd the wor ld assum i ng tha t it p rov i de s at lea s t some re l a t i ve i n f orma shyt i on Wh i le i n the U n i ted States re a l pe r cap i t a pe r s ona l i ncome grew by 5 3 pe r ce n t from 19 5 0 - 1 9 7 0 a compa r a b le we i gh te d shya ve ra ge income f i gu re for the wor ld grew by 1 38 pe rce nt in the s ame pe r i od whe n the we i gh t s used re f le c t the actu a l d i s t r i bu t i on o f FMG imm i g r a n t s in 1 9 68 imme d i a te l y pr i or to enactme nt of the new imm i g rat i on law Moreove r in a 1 9 68 su rve y the ra t i o of p hy s i c i a n to pe rs ona l income was hi gher in As i an count r i e s th an i n the U S equa l i ng 16 9 i n the Ph i l i pp i ne s 34 9 in Tha i la n d a nd 8 5 i n the us acco rd i ng t o Me j i a et al sup ra note 12 a t 1 3 1 The re f ore i t seems l ike l y th at the-oppor t u n i ty cos t for f ore i gn me d i ca l graduates rose du r i ng the pe r i od in que s t i on I t i s true tha t the abo l i t i on of the na t i ona l quo t a sy s tem i n the Imm i gr a t i on Act in 1 9 68 se rve d to ope n the doo r s to phy s i c i ans f rom poorer cou n t r i e s and he nce w i th lowe r oppor tu n i ty cos t s The con com i t ant exp ans i on in em i gr a t i on of As i a n FMG1s howe ve r i s the re s u l t of ea s i ng re gu l a t i on not a lowe r i ng of the i r opportu n i ty cos t

- 1 4 shy

w i l l be se t to 1 af te r 1 9 7 6 (D7 6 ) 1 7 I f re s t r i c t i ons are ef fec shy

t i ve bo th dummy va r i a b le s shou ld have neg a t i ve coe f f i c ie n t s

Three l i ce n s i ng va r i a b le s are used One mea s u res the ef fec t

o f states th a t d i d not l i ce nse FMG s unde r any c i rcum s t ance s

( NOL IC ) the se cond takes ac count o f statutes w i th c i t i ze ns h i p

requ i reme n t s (C I T I Z ) and the th i rd es t imates the impact of

gra du ate tra i n i ng t ime requ i reme n ts (TNGTM ) A l l l i ce n s i n g

v a r i ables are me a s u re d a s the we i gh ted pe rce nt age o f a l l states

w i t h the re le va nt re s t r i c t i on in ea ch yea r 18 F i gure 4 shows the

p a t te rn of the se th ree va r i a b les ove r t ime The leve l s of NOL I C

a nd C I TIZ are h i ghes t in the 1 9 5 0 s and su bsequent l y de c l i ne

the i r re l a xa t i on may exp l a i n some of the gradu a l incre a se in FMG

e nt ry du r i n g the 1 9 5 0 s and ea r ly 1 9 6 0 s TNGTM excep t for a

d i p in the 1 9 7 0 s has r i se n

The FMG -popu l a t i on ra t i o (FMGP ) is used as a de pende nt

var i ab le 1 9 Bec a u se of ava i l a b le da t a the f r i nge is mea s u re d as

17 A dummy va r i a b le equ a l to 1 be f ore 1 9 68 was su bs t i t u ted for D 7 1 in some re gre s s i on s I t was ge ne ra l l y not s i gn i f i c ant p re sumab l y e i the r be c a u se the 1 9 7 1 law cha n ge was more import ant or be cause the FMG ent ry re s p onse oc c u r re d with a lag These res u l t are not repor ted here See Noe ther sup r a note 1 1 for them

18 S i nce the m i d - 1 9 6 0 s the we i gh ts used a re the pe rce n tage of nat i ona l pe r s ona l hea l t h ca re expe nd i tu re s con t r i bu ted by ea ch s ta te P r i o r to that t ime st a te bre ak downs of hea l t h expe nd i shyt u res are not ava i l a b le so pe r s on a l i ncome i s used as a we i gh t i ng f ac to r i n s tead U s i n g i ncome as the we i ght i ng factor th rou ghout p rodu ce s s im i l a r re s u l t s

1 9 Regre s s i on s we re a l s o ru n on the leve l numbe r of FMG s as we l l a s the FMG-new l i censee s ra t i o Resu l t s d i d not di f fe r s i gn i f i c a n t l y from those reported

- 1 5 shy

bull bull bull bull

bull bull bull

bull bull

bull bull bull bull bull bull

50

bull bull bull

bull bull

FIGURE 4

FMG Re s t r i c t i on s

per ent (a) FMG licensing (+) and citizenship () requirements

I i

amp0 bull

I I I

bull bull

bull bull bull

lO

bull

o bull

10 bull bull

bull -

bull6 amp4 50 SZ bull 3i 51 iiO iZ Si ampe 70 z 7bull 7amp 79 80

YEAR

(b) FMG graduate training time requirementsyears 1 i

II

1 5 I 1

bull

bull w bull bull

bull 1 bull bull

bull o

bullCll

e

i I

01 - bull

c a I bull -cs

a f low of new ent r a n t s ( l i ce nsees ) ra the r th a n as the stock

e x i s t i ng each yea r By cumu l a t i ng the yea r l y f low ma gn i t udes

ove r the pe r i od i n que s t i on we ca n es t imate the tot a l impa c t of

the f r i nge on us phy s i c i an supp l y The FMG ent ry equ a t i on to

be es t imated i s

A

( 4 ) FMGP t = fQ + f l MD I NC t + f 2 1 D7 l t + f 2 2 D 7 6 t + f 2 3 NOL I C t

us phy s i c i a n income is not a t rue exoge nou s va r i a b le s i n ce

i t is obv i ou s ly af f e c ted by the numbe r of FMG s th at enter the

U n i ted St a te s The re f ore two -s t a ge lea s t squ a re s is used to

e s t imate the sy s tem of ( 4 ) and a reduced form income equ a t i on to

de s c r i be MD I NC th at con t a i ns va r i ables to me a s u re both fore i g n

a nd dome s t i c ent ry re s t r i c t i o n s as we l l as dema nd parame te r s and

that ac c o u nt s f or gradu a l supp ly ad j u s tme n t s to equ i l i br i um

c hange s th rou gh i n c l u s i on of the lagged stock and a pre d i c t i on

e rror te rm

2 0 I n the es t i ma t i on no va r i a b le is used to account for FMG oppor tu n i ty cos t S i nce fore i gn phy s i c i a n income da t a are not rea d i ly ava i l a b le interna t i on a l income da t a we re used to crea te a proxy f or FMGTOC A we i ghted ave ra ge of va r i ou s wor l d a re a s (eg Eas te r n As i a ) pe r cap i t a gros s na t i ona l produ c t in con s t a n t do l l a r s was ca l c u l a te d us i ng the 1 9 68 na t i ona l propor shyt i ons of FMG en t r a nts as we i gh t s A s im i la r va r i a b le was a l s o t r ied us i ng 1 9 6 5 we i g h ts B o t h showe d i na p p rop r i ate pos i t i ve e f f e c t s on FMG ent ry and proba b l y re f le cted factors not re l a ted t o FMG opp ortu n i ty cos t of imm i g r a t i n g to the us

- 1 7 shy

This re duced form income equation is de rive d from demand and

c o s t equa tions the unde r lying mode l for which is des c ribe d in

g re a te r de t ai l in Appe ndi x A I t uses pe r capita rea l in come

( INCM ) and the pe rce n t of the popu l a tion co ve re d by pu b lic insu r shy

a nce (PUB I NS ) as dem a n d va riable s 2 1 They shou l d both have a

positive e f fect on phy sician income Margina l cos t is me a s u red

as the tot a l opp ortunity cos t (TOC ) of be coming a physicia n

based on the as s ump tion tha t the ne xt be s t a l te rna tive is workin g

f o l lowin g fou r yea r s of co l le ge a n d ad j u s ted f o r the time spe nt

2 2in me di c a l schoo l and re side n cy prog r ams It shou l d a l s o be

p ositive l y co rre l a te d wit h phy sician income V aria b le s tha t

mea s u re fede ra l subsidies gra nted t o us medica l schoo l s bo th

f or con s t ru ction (CON S 8 ) and as capitation grants (CAPS 5 ) a re

a ls o in corporated and wil l be dis c u s se d fu rthe r in Section I I I

be low Whi le a true re s t ric tion on a dome s tic f ringe shou ld

r aise phy sician income since these are subsidie s they ease

e n t ry of the dome s tic f ringe and he nce lowe r phy sician in come

Fina l ly phy sician income is a f fe c te d by di f fe re n ce s be twee n

a c tu a l ent ry int o the profession and th at pre dicted by the

dominant firm As a re lated pape r shows this error ca n be

2 l O the r demand va ri a b le s we re a l s o used in Noe t he r sup r a notes 1 and 1 1 and these appe a re d mos t signific a n t Hence othe r variable s su ch as edu ca tion and age we re omit ted to avoid mu lti-co l line a rity prob lems with other independe n t va ri a b le s

2 2 The de riva tion and me a s u reme nt of TOC are fu l ly de s c ribe d in N oe t he r supra note 1 1 D at a sou rce s for TOC and the other v a ria bles used are lis ted in Appendi x B

- 1 8shy

supra

re duced to incorrect pre d i ct i ons about FMG ent ry wh i ch we de note

e F 2 3 Thu s the est i mated equ at i on i s as fol lows

( 5 ) MDINCt = oo + 0 1 TOCt + 0 2 1 INCMt + 0 2 2 PUBINSt + 0 3 1 D7 1t

+ 03 2 D7 6t + 0 3 3 NOLICt + 0 3 4 CITI Zt + 0 4 1 CON S 8t

F 2 + 04 2 CAPS 5t + 0 5 MDSPt - 1 + 0 6 et + Et

Add i t i ona l ly the re du c e d form equat i on de s cr i b i ng FMG

e nt ry de r i ved by su bst itut i ng ( 5 ) i nto ( 4 ) is a l s o est imated

a s

( 6 ) F MGPt = + TOCt + 2 1 INCMt + 2 2 PUBINSt + 3 1 D7 1t0 1

+ 3 2 D7 6t + 3 3 NOLICt + 3 4 CITI Zt + 3 5 TNGTMt

F 3 + 4 1 CON S 8t + 4 2 CAPS 5t + 5 MDSPt - 1 + 6 et + Et

The ef f e cts of the FMG re str i ct i on s and opp o rtu n i ty cost on the

redu c e d form of FMG are i ndete rm i nate The i r d i r e ct ef f e ct is

obv i ou s ly ne gat i ve but th ey have a pos it i ve i n f l u e n c e on i ncome

w h i ch a l s o a f f e cts the dep e nd e nt va r i a b l e pos it i ve l y P re s umshy

a b l y the forme r ef f e ct is strong e r so that the net impact is

negat i ve The rema i n i ng va r i a b l e s have th e same i nf luence as

they do on MDINC i e th e coe f f i c i e nts on TOC INCM PUBINS

2 3 eFSee Noethe r note 1 1 The me asu reme nt of is d e s cr i be d in Append i x B

- 1 9 shy

FCON S 8 and CAPS 5 are pos it i ve wh i l e tho s e on MDS Pt - 1 and e are

negat i ve

Column s 1 and 2 of tab l e 1 pre s e nt the re s u lts from a 2SLS

regre s s i on of the sy stem conta i n i ng the re duc e d form income

equat i on ( 5 ) and th e FMG supp ly cu rve ( 4 ) In the MDINC equat i on

a l l the fore i gn and dome st i c re str i ct i ons show the pre d i ct e d

s i gns t h e FMG ent ry re st r i ct i ons inc r e a s e phy s i c i a n income

wh i le subs i d i es to us me d i c a l schoo l s de c r e a s e it The

p re d i ct i on of phy s i c i a n i ncome de r i ved here is used as the MDINC

va r i a b l e i n the FMG equ at i on s reported here in co lumn 2 It has

the expe cted pos it i ve ef f e ct A l l of the re str i ct i on va r i a b l e s

have the ant i c i p at e d negat i ve e f f e ct on FMG ent ry 2 4

Wh i l e th e re l a xat i on o f l i ce n s i ng and imm i grat i on laws

app arent ly d i d sh i ft the FMG supp l y cu rve to the r i g ht the

coe f f i c i e nt on the D7 6 va r i a b l e a l s o su ggests th at the cu rve has

s h i fted ba c k in rec e nt ye a rs In f a ct the coe f f i c i e nt on D7 6 i s

a lmost a s large a s that on D7 1 F rom th e est i mated FMG equ at i on s

and the pre d i ct i ons o f MDINC we c a n tra ce th e moveme nt of the

F MG suppl y cu rve ove r time F i gure 5 tra c e s its moveme nt It

s hows that afte r rema i n i ng re l at i ve ly stat i o n a ry unt i l th e lat e

1 9 60 s the supp l y sh i fted out subst a nt i a l ly i n the f i rst ha l f of

the 1 9 70 s The outw a rd moveme nt cont i nu e d unt i l about 1 9 7 5 du e

2 4 E ither i n c l u d i ng TNGTM i n the i ncome equ at i on or not i nc l u d i ng it in the FMG equat i on doe s not ch ange the re su lt s s i g n i f i c a nt ly The s imp le co rre lat i on betwe e n TOC and TNGTM e xc e e d s 9

- 20 shy

w i l l be se t to 1 af te r 1 9 7 6 (D7 6 ) 1 7 I f re s t r i c t i ons are ef fec shy

t i ve bo th dummy va r i a b le s shou ld have neg a t i ve coe f f i c ie n t s

Three l i ce n s i ng va r i a b le s are used One mea s u res the ef fec t

o f states th a t d i d not l i ce nse FMG s unde r any c i rcum s t ance s

( NOL IC ) the se cond takes ac count o f statutes w i th c i t i ze ns h i p

requ i reme n t s (C I T I Z ) and the th i rd es t imates the impact of

gra du ate tra i n i ng t ime requ i reme n ts (TNGTM ) A l l l i ce n s i n g

v a r i ables are me a s u re d a s the we i gh ted pe rce nt age o f a l l states

w i t h the re le va nt re s t r i c t i on in ea ch yea r 18 F i gure 4 shows the

p a t te rn of the se th ree va r i a b les ove r t ime The leve l s of NOL I C

a nd C I TIZ are h i ghes t in the 1 9 5 0 s and su bsequent l y de c l i ne

the i r re l a xa t i on may exp l a i n some of the gradu a l incre a se in FMG

e nt ry du r i n g the 1 9 5 0 s and ea r ly 1 9 6 0 s TNGTM excep t for a

d i p in the 1 9 7 0 s has r i se n

The FMG -popu l a t i on ra t i o (FMGP ) is used as a de pende nt

var i ab le 1 9 Bec a u se of ava i l a b le da t a the f r i nge is mea s u re d as

17 A dummy va r i a b le equ a l to 1 be f ore 1 9 68 was su bs t i t u ted for D 7 1 in some re gre s s i on s I t was ge ne ra l l y not s i gn i f i c ant p re sumab l y e i the r be c a u se the 1 9 7 1 law cha n ge was more import ant or be cause the FMG ent ry re s p onse oc c u r re d with a lag These res u l t are not repor ted here See Noe ther sup r a note 1 1 for them

18 S i nce the m i d - 1 9 6 0 s the we i gh ts used a re the pe rce n tage of nat i ona l pe r s ona l hea l t h ca re expe nd i tu re s con t r i bu ted by ea ch s ta te P r i o r to that t ime st a te bre ak downs of hea l t h expe nd i shyt u res are not ava i l a b le so pe r s on a l i ncome i s used as a we i gh t i ng f ac to r i n s tead U s i n g i ncome as the we i ght i ng factor th rou ghout p rodu ce s s im i l a r re s u l t s

1 9 Regre s s i on s we re a l s o ru n on the leve l numbe r of FMG s as we l l a s the FMG-new l i censee s ra t i o Resu l t s d i d not di f fe r s i gn i f i c a n t l y from those reported

- 1 5 shy

bull bull bull bull

bull bull bull

bull bull

bull bull bull bull bull bull

50

bull bull bull

bull bull

FIGURE 4

FMG Re s t r i c t i on s

per ent (a) FMG licensing (+) and citizenship () requirements

I i

amp0 bull

I I I

bull bull

bull bull bull

lO

bull

o bull

10 bull bull

bull -

bull6 amp4 50 SZ bull 3i 51 iiO iZ Si ampe 70 z 7bull 7amp 79 80

YEAR

(b) FMG graduate training time requirementsyears 1 i

II

1 5 I 1

bull

bull w bull bull

bull 1 bull bull

bull o

bullCll

e

i I

01 - bull

c a I bull -cs

a f low of new ent r a n t s ( l i ce nsees ) ra the r th a n as the stock

e x i s t i ng each yea r By cumu l a t i ng the yea r l y f low ma gn i t udes

ove r the pe r i od i n que s t i on we ca n es t imate the tot a l impa c t of

the f r i nge on us phy s i c i an supp l y The FMG ent ry equ a t i on to

be es t imated i s

A

( 4 ) FMGP t = fQ + f l MD I NC t + f 2 1 D7 l t + f 2 2 D 7 6 t + f 2 3 NOL I C t

us phy s i c i a n income is not a t rue exoge nou s va r i a b le s i n ce

i t is obv i ou s ly af f e c ted by the numbe r of FMG s th at enter the

U n i ted St a te s The re f ore two -s t a ge lea s t squ a re s is used to

e s t imate the sy s tem of ( 4 ) and a reduced form income equ a t i on to

de s c r i be MD I NC th at con t a i ns va r i ables to me a s u re both fore i g n

a nd dome s t i c ent ry re s t r i c t i o n s as we l l as dema nd parame te r s and

that ac c o u nt s f or gradu a l supp ly ad j u s tme n t s to equ i l i br i um

c hange s th rou gh i n c l u s i on of the lagged stock and a pre d i c t i on

e rror te rm

2 0 I n the es t i ma t i on no va r i a b le is used to account for FMG oppor tu n i ty cos t S i nce fore i gn phy s i c i a n income da t a are not rea d i ly ava i l a b le interna t i on a l income da t a we re used to crea te a proxy f or FMGTOC A we i ghted ave ra ge of va r i ou s wor l d a re a s (eg Eas te r n As i a ) pe r cap i t a gros s na t i ona l produ c t in con s t a n t do l l a r s was ca l c u l a te d us i ng the 1 9 68 na t i ona l propor shyt i ons of FMG en t r a nts as we i gh t s A s im i la r va r i a b le was a l s o t r ied us i ng 1 9 6 5 we i g h ts B o t h showe d i na p p rop r i ate pos i t i ve e f f e c t s on FMG ent ry and proba b l y re f le cted factors not re l a ted t o FMG opp ortu n i ty cos t of imm i g r a t i n g to the us

- 1 7 shy

This re duced form income equation is de rive d from demand and

c o s t equa tions the unde r lying mode l for which is des c ribe d in

g re a te r de t ai l in Appe ndi x A I t uses pe r capita rea l in come

( INCM ) and the pe rce n t of the popu l a tion co ve re d by pu b lic insu r shy

a nce (PUB I NS ) as dem a n d va riable s 2 1 They shou l d both have a

positive e f fect on phy sician income Margina l cos t is me a s u red

as the tot a l opp ortunity cos t (TOC ) of be coming a physicia n

based on the as s ump tion tha t the ne xt be s t a l te rna tive is workin g

f o l lowin g fou r yea r s of co l le ge a n d ad j u s ted f o r the time spe nt

2 2in me di c a l schoo l and re side n cy prog r ams It shou l d a l s o be

p ositive l y co rre l a te d wit h phy sician income V aria b le s tha t

mea s u re fede ra l subsidies gra nted t o us medica l schoo l s bo th

f or con s t ru ction (CON S 8 ) and as capitation grants (CAPS 5 ) a re

a ls o in corporated and wil l be dis c u s se d fu rthe r in Section I I I

be low Whi le a true re s t ric tion on a dome s tic f ringe shou ld

r aise phy sician income since these are subsidie s they ease

e n t ry of the dome s tic f ringe and he nce lowe r phy sician in come

Fina l ly phy sician income is a f fe c te d by di f fe re n ce s be twee n

a c tu a l ent ry int o the profession and th at pre dicted by the

dominant firm As a re lated pape r shows this error ca n be

2 l O the r demand va ri a b le s we re a l s o used in Noe t he r sup r a notes 1 and 1 1 and these appe a re d mos t signific a n t Hence othe r variable s su ch as edu ca tion and age we re omit ted to avoid mu lti-co l line a rity prob lems with other independe n t va ri a b le s

2 2 The de riva tion and me a s u reme nt of TOC are fu l ly de s c ribe d in N oe t he r supra note 1 1 D at a sou rce s for TOC and the other v a ria bles used are lis ted in Appendi x B

- 1 8shy

supra

re duced to incorrect pre d i ct i ons about FMG ent ry wh i ch we de note

e F 2 3 Thu s the est i mated equ at i on i s as fol lows

( 5 ) MDINCt = oo + 0 1 TOCt + 0 2 1 INCMt + 0 2 2 PUBINSt + 0 3 1 D7 1t

+ 03 2 D7 6t + 0 3 3 NOLICt + 0 3 4 CITI Zt + 0 4 1 CON S 8t

F 2 + 04 2 CAPS 5t + 0 5 MDSPt - 1 + 0 6 et + Et

Add i t i ona l ly the re du c e d form equat i on de s cr i b i ng FMG

e nt ry de r i ved by su bst itut i ng ( 5 ) i nto ( 4 ) is a l s o est imated

a s

( 6 ) F MGPt = + TOCt + 2 1 INCMt + 2 2 PUBINSt + 3 1 D7 1t0 1

+ 3 2 D7 6t + 3 3 NOLICt + 3 4 CITI Zt + 3 5 TNGTMt

F 3 + 4 1 CON S 8t + 4 2 CAPS 5t + 5 MDSPt - 1 + 6 et + Et

The ef f e cts of the FMG re str i ct i on s and opp o rtu n i ty cost on the

redu c e d form of FMG are i ndete rm i nate The i r d i r e ct ef f e ct is

obv i ou s ly ne gat i ve but th ey have a pos it i ve i n f l u e n c e on i ncome

w h i ch a l s o a f f e cts the dep e nd e nt va r i a b l e pos it i ve l y P re s umshy

a b l y the forme r ef f e ct is strong e r so that the net impact is

negat i ve The rema i n i ng va r i a b l e s have th e same i nf luence as

they do on MDINC i e th e coe f f i c i e nts on TOC INCM PUBINS

2 3 eFSee Noethe r note 1 1 The me asu reme nt of is d e s cr i be d in Append i x B

- 1 9 shy

FCON S 8 and CAPS 5 are pos it i ve wh i l e tho s e on MDS Pt - 1 and e are

negat i ve

Column s 1 and 2 of tab l e 1 pre s e nt the re s u lts from a 2SLS

regre s s i on of the sy stem conta i n i ng the re duc e d form income

equat i on ( 5 ) and th e FMG supp ly cu rve ( 4 ) In the MDINC equat i on

a l l the fore i gn and dome st i c re str i ct i ons show the pre d i ct e d

s i gns t h e FMG ent ry re st r i ct i ons inc r e a s e phy s i c i a n income

wh i le subs i d i es to us me d i c a l schoo l s de c r e a s e it The

p re d i ct i on of phy s i c i a n i ncome de r i ved here is used as the MDINC

va r i a b l e i n the FMG equ at i on s reported here in co lumn 2 It has

the expe cted pos it i ve ef f e ct A l l of the re str i ct i on va r i a b l e s

have the ant i c i p at e d negat i ve e f f e ct on FMG ent ry 2 4

Wh i l e th e re l a xat i on o f l i ce n s i ng and imm i grat i on laws

app arent ly d i d sh i ft the FMG supp l y cu rve to the r i g ht the

coe f f i c i e nt on the D7 6 va r i a b l e a l s o su ggests th at the cu rve has

s h i fted ba c k in rec e nt ye a rs In f a ct the coe f f i c i e nt on D7 6 i s

a lmost a s large a s that on D7 1 F rom th e est i mated FMG equ at i on s

and the pre d i ct i ons o f MDINC we c a n tra ce th e moveme nt of the

F MG suppl y cu rve ove r time F i gure 5 tra c e s its moveme nt It

s hows that afte r rema i n i ng re l at i ve ly stat i o n a ry unt i l th e lat e

1 9 60 s the supp l y sh i fted out subst a nt i a l ly i n the f i rst ha l f of

the 1 9 70 s The outw a rd moveme nt cont i nu e d unt i l about 1 9 7 5 du e

2 4 E ither i n c l u d i ng TNGTM i n the i ncome equ at i on or not i nc l u d i ng it in the FMG equat i on doe s not ch ange the re su lt s s i g n i f i c a nt ly The s imp le co rre lat i on betwe e n TOC and TNGTM e xc e e d s 9

- 20 shy

bull bull bull bull

bull bull bull

bull bull

bull bull bull bull bull bull

50

bull bull bull

bull bull

FIGURE 4

FMG Re s t r i c t i on s

per ent (a) FMG licensing (+) and citizenship () requirements

I i

amp0 bull

I I I

bull bull

bull bull bull

lO

bull

o bull

10 bull bull

bull -

bull6 amp4 50 SZ bull 3i 51 iiO iZ Si ampe 70 z 7bull 7amp 79 80

YEAR

(b) FMG graduate training time requirementsyears 1 i

II

1 5 I 1

bull

bull w bull bull

bull 1 bull bull

bull o

bullCll

e

i I

01 - bull

c a I bull -cs

a f low of new ent r a n t s ( l i ce nsees ) ra the r th a n as the stock

e x i s t i ng each yea r By cumu l a t i ng the yea r l y f low ma gn i t udes

ove r the pe r i od i n que s t i on we ca n es t imate the tot a l impa c t of

the f r i nge on us phy s i c i an supp l y The FMG ent ry equ a t i on to

be es t imated i s

A

( 4 ) FMGP t = fQ + f l MD I NC t + f 2 1 D7 l t + f 2 2 D 7 6 t + f 2 3 NOL I C t

us phy s i c i a n income is not a t rue exoge nou s va r i a b le s i n ce

i t is obv i ou s ly af f e c ted by the numbe r of FMG s th at enter the

U n i ted St a te s The re f ore two -s t a ge lea s t squ a re s is used to

e s t imate the sy s tem of ( 4 ) and a reduced form income equ a t i on to

de s c r i be MD I NC th at con t a i ns va r i ables to me a s u re both fore i g n

a nd dome s t i c ent ry re s t r i c t i o n s as we l l as dema nd parame te r s and

that ac c o u nt s f or gradu a l supp ly ad j u s tme n t s to equ i l i br i um

c hange s th rou gh i n c l u s i on of the lagged stock and a pre d i c t i on

e rror te rm

2 0 I n the es t i ma t i on no va r i a b le is used to account for FMG oppor tu n i ty cos t S i nce fore i gn phy s i c i a n income da t a are not rea d i ly ava i l a b le interna t i on a l income da t a we re used to crea te a proxy f or FMGTOC A we i ghted ave ra ge of va r i ou s wor l d a re a s (eg Eas te r n As i a ) pe r cap i t a gros s na t i ona l produ c t in con s t a n t do l l a r s was ca l c u l a te d us i ng the 1 9 68 na t i ona l propor shyt i ons of FMG en t r a nts as we i gh t s A s im i la r va r i a b le was a l s o t r ied us i ng 1 9 6 5 we i g h ts B o t h showe d i na p p rop r i ate pos i t i ve e f f e c t s on FMG ent ry and proba b l y re f le cted factors not re l a ted t o FMG opp ortu n i ty cos t of imm i g r a t i n g to the us

- 1 7 shy

This re duced form income equation is de rive d from demand and

c o s t equa tions the unde r lying mode l for which is des c ribe d in

g re a te r de t ai l in Appe ndi x A I t uses pe r capita rea l in come

( INCM ) and the pe rce n t of the popu l a tion co ve re d by pu b lic insu r shy

a nce (PUB I NS ) as dem a n d va riable s 2 1 They shou l d both have a

positive e f fect on phy sician income Margina l cos t is me a s u red

as the tot a l opp ortunity cos t (TOC ) of be coming a physicia n

based on the as s ump tion tha t the ne xt be s t a l te rna tive is workin g

f o l lowin g fou r yea r s of co l le ge a n d ad j u s ted f o r the time spe nt

2 2in me di c a l schoo l and re side n cy prog r ams It shou l d a l s o be

p ositive l y co rre l a te d wit h phy sician income V aria b le s tha t

mea s u re fede ra l subsidies gra nted t o us medica l schoo l s bo th

f or con s t ru ction (CON S 8 ) and as capitation grants (CAPS 5 ) a re

a ls o in corporated and wil l be dis c u s se d fu rthe r in Section I I I

be low Whi le a true re s t ric tion on a dome s tic f ringe shou ld

r aise phy sician income since these are subsidie s they ease

e n t ry of the dome s tic f ringe and he nce lowe r phy sician in come

Fina l ly phy sician income is a f fe c te d by di f fe re n ce s be twee n

a c tu a l ent ry int o the profession and th at pre dicted by the

dominant firm As a re lated pape r shows this error ca n be

2 l O the r demand va ri a b le s we re a l s o used in Noe t he r sup r a notes 1 and 1 1 and these appe a re d mos t signific a n t Hence othe r variable s su ch as edu ca tion and age we re omit ted to avoid mu lti-co l line a rity prob lems with other independe n t va ri a b le s

2 2 The de riva tion and me a s u reme nt of TOC are fu l ly de s c ribe d in N oe t he r supra note 1 1 D at a sou rce s for TOC and the other v a ria bles used are lis ted in Appendi x B

- 1 8shy

supra

re duced to incorrect pre d i ct i ons about FMG ent ry wh i ch we de note

e F 2 3 Thu s the est i mated equ at i on i s as fol lows

( 5 ) MDINCt = oo + 0 1 TOCt + 0 2 1 INCMt + 0 2 2 PUBINSt + 0 3 1 D7 1t

+ 03 2 D7 6t + 0 3 3 NOLICt + 0 3 4 CITI Zt + 0 4 1 CON S 8t

F 2 + 04 2 CAPS 5t + 0 5 MDSPt - 1 + 0 6 et + Et

Add i t i ona l ly the re du c e d form equat i on de s cr i b i ng FMG

e nt ry de r i ved by su bst itut i ng ( 5 ) i nto ( 4 ) is a l s o est imated

a s

( 6 ) F MGPt = + TOCt + 2 1 INCMt + 2 2 PUBINSt + 3 1 D7 1t0 1

+ 3 2 D7 6t + 3 3 NOLICt + 3 4 CITI Zt + 3 5 TNGTMt

F 3 + 4 1 CON S 8t + 4 2 CAPS 5t + 5 MDSPt - 1 + 6 et + Et

The ef f e cts of the FMG re str i ct i on s and opp o rtu n i ty cost on the

redu c e d form of FMG are i ndete rm i nate The i r d i r e ct ef f e ct is

obv i ou s ly ne gat i ve but th ey have a pos it i ve i n f l u e n c e on i ncome

w h i ch a l s o a f f e cts the dep e nd e nt va r i a b l e pos it i ve l y P re s umshy

a b l y the forme r ef f e ct is strong e r so that the net impact is

negat i ve The rema i n i ng va r i a b l e s have th e same i nf luence as

they do on MDINC i e th e coe f f i c i e nts on TOC INCM PUBINS

2 3 eFSee Noethe r note 1 1 The me asu reme nt of is d e s cr i be d in Append i x B

- 1 9 shy

FCON S 8 and CAPS 5 are pos it i ve wh i l e tho s e on MDS Pt - 1 and e are

negat i ve

Column s 1 and 2 of tab l e 1 pre s e nt the re s u lts from a 2SLS

regre s s i on of the sy stem conta i n i ng the re duc e d form income

equat i on ( 5 ) and th e FMG supp ly cu rve ( 4 ) In the MDINC equat i on

a l l the fore i gn and dome st i c re str i ct i ons show the pre d i ct e d

s i gns t h e FMG ent ry re st r i ct i ons inc r e a s e phy s i c i a n income

wh i le subs i d i es to us me d i c a l schoo l s de c r e a s e it The

p re d i ct i on of phy s i c i a n i ncome de r i ved here is used as the MDINC

va r i a b l e i n the FMG equ at i on s reported here in co lumn 2 It has

the expe cted pos it i ve ef f e ct A l l of the re str i ct i on va r i a b l e s

have the ant i c i p at e d negat i ve e f f e ct on FMG ent ry 2 4

Wh i l e th e re l a xat i on o f l i ce n s i ng and imm i grat i on laws

app arent ly d i d sh i ft the FMG supp l y cu rve to the r i g ht the

coe f f i c i e nt on the D7 6 va r i a b l e a l s o su ggests th at the cu rve has

s h i fted ba c k in rec e nt ye a rs In f a ct the coe f f i c i e nt on D7 6 i s

a lmost a s large a s that on D7 1 F rom th e est i mated FMG equ at i on s

and the pre d i ct i ons o f MDINC we c a n tra ce th e moveme nt of the

F MG suppl y cu rve ove r time F i gure 5 tra c e s its moveme nt It

s hows that afte r rema i n i ng re l at i ve ly stat i o n a ry unt i l th e lat e

1 9 60 s the supp l y sh i fted out subst a nt i a l ly i n the f i rst ha l f of

the 1 9 70 s The outw a rd moveme nt cont i nu e d unt i l about 1 9 7 5 du e

2 4 E ither i n c l u d i ng TNGTM i n the i ncome equ at i on or not i nc l u d i ng it in the FMG equat i on doe s not ch ange the re su lt s s i g n i f i c a nt ly The s imp le co rre lat i on betwe e n TOC and TNGTM e xc e e d s 9

- 20 shy

a f low of new ent r a n t s ( l i ce nsees ) ra the r th a n as the stock

e x i s t i ng each yea r By cumu l a t i ng the yea r l y f low ma gn i t udes

ove r the pe r i od i n que s t i on we ca n es t imate the tot a l impa c t of

the f r i nge on us phy s i c i an supp l y The FMG ent ry equ a t i on to

be es t imated i s

A

( 4 ) FMGP t = fQ + f l MD I NC t + f 2 1 D7 l t + f 2 2 D 7 6 t + f 2 3 NOL I C t

us phy s i c i a n income is not a t rue exoge nou s va r i a b le s i n ce

i t is obv i ou s ly af f e c ted by the numbe r of FMG s th at enter the

U n i ted St a te s The re f ore two -s t a ge lea s t squ a re s is used to

e s t imate the sy s tem of ( 4 ) and a reduced form income equ a t i on to

de s c r i be MD I NC th at con t a i ns va r i ables to me a s u re both fore i g n

a nd dome s t i c ent ry re s t r i c t i o n s as we l l as dema nd parame te r s and

that ac c o u nt s f or gradu a l supp ly ad j u s tme n t s to equ i l i br i um

c hange s th rou gh i n c l u s i on of the lagged stock and a pre d i c t i on

e rror te rm

2 0 I n the es t i ma t i on no va r i a b le is used to account for FMG oppor tu n i ty cos t S i nce fore i gn phy s i c i a n income da t a are not rea d i ly ava i l a b le interna t i on a l income da t a we re used to crea te a proxy f or FMGTOC A we i ghted ave ra ge of va r i ou s wor l d a re a s (eg Eas te r n As i a ) pe r cap i t a gros s na t i ona l produ c t in con s t a n t do l l a r s was ca l c u l a te d us i ng the 1 9 68 na t i ona l propor shyt i ons of FMG en t r a nts as we i gh t s A s im i la r va r i a b le was a l s o t r ied us i ng 1 9 6 5 we i g h ts B o t h showe d i na p p rop r i ate pos i t i ve e f f e c t s on FMG ent ry and proba b l y re f le cted factors not re l a ted t o FMG opp ortu n i ty cos t of imm i g r a t i n g to the us

- 1 7 shy

This re duced form income equation is de rive d from demand and

c o s t equa tions the unde r lying mode l for which is des c ribe d in

g re a te r de t ai l in Appe ndi x A I t uses pe r capita rea l in come

( INCM ) and the pe rce n t of the popu l a tion co ve re d by pu b lic insu r shy

a nce (PUB I NS ) as dem a n d va riable s 2 1 They shou l d both have a

positive e f fect on phy sician income Margina l cos t is me a s u red

as the tot a l opp ortunity cos t (TOC ) of be coming a physicia n

based on the as s ump tion tha t the ne xt be s t a l te rna tive is workin g

f o l lowin g fou r yea r s of co l le ge a n d ad j u s ted f o r the time spe nt

2 2in me di c a l schoo l and re side n cy prog r ams It shou l d a l s o be

p ositive l y co rre l a te d wit h phy sician income V aria b le s tha t

mea s u re fede ra l subsidies gra nted t o us medica l schoo l s bo th

f or con s t ru ction (CON S 8 ) and as capitation grants (CAPS 5 ) a re

a ls o in corporated and wil l be dis c u s se d fu rthe r in Section I I I

be low Whi le a true re s t ric tion on a dome s tic f ringe shou ld

r aise phy sician income since these are subsidie s they ease

e n t ry of the dome s tic f ringe and he nce lowe r phy sician in come

Fina l ly phy sician income is a f fe c te d by di f fe re n ce s be twee n

a c tu a l ent ry int o the profession and th at pre dicted by the

dominant firm As a re lated pape r shows this error ca n be

2 l O the r demand va ri a b le s we re a l s o used in Noe t he r sup r a notes 1 and 1 1 and these appe a re d mos t signific a n t Hence othe r variable s su ch as edu ca tion and age we re omit ted to avoid mu lti-co l line a rity prob lems with other independe n t va ri a b le s

2 2 The de riva tion and me a s u reme nt of TOC are fu l ly de s c ribe d in N oe t he r supra note 1 1 D at a sou rce s for TOC and the other v a ria bles used are lis ted in Appendi x B

- 1 8shy

supra

re duced to incorrect pre d i ct i ons about FMG ent ry wh i ch we de note

e F 2 3 Thu s the est i mated equ at i on i s as fol lows

( 5 ) MDINCt = oo + 0 1 TOCt + 0 2 1 INCMt + 0 2 2 PUBINSt + 0 3 1 D7 1t

+ 03 2 D7 6t + 0 3 3 NOLICt + 0 3 4 CITI Zt + 0 4 1 CON S 8t

F 2 + 04 2 CAPS 5t + 0 5 MDSPt - 1 + 0 6 et + Et

Add i t i ona l ly the re du c e d form equat i on de s cr i b i ng FMG

e nt ry de r i ved by su bst itut i ng ( 5 ) i nto ( 4 ) is a l s o est imated

a s

( 6 ) F MGPt = + TOCt + 2 1 INCMt + 2 2 PUBINSt + 3 1 D7 1t0 1

+ 3 2 D7 6t + 3 3 NOLICt + 3 4 CITI Zt + 3 5 TNGTMt

F 3 + 4 1 CON S 8t + 4 2 CAPS 5t + 5 MDSPt - 1 + 6 et + Et

The ef f e cts of the FMG re str i ct i on s and opp o rtu n i ty cost on the

redu c e d form of FMG are i ndete rm i nate The i r d i r e ct ef f e ct is

obv i ou s ly ne gat i ve but th ey have a pos it i ve i n f l u e n c e on i ncome

w h i ch a l s o a f f e cts the dep e nd e nt va r i a b l e pos it i ve l y P re s umshy

a b l y the forme r ef f e ct is strong e r so that the net impact is

negat i ve The rema i n i ng va r i a b l e s have th e same i nf luence as

they do on MDINC i e th e coe f f i c i e nts on TOC INCM PUBINS

2 3 eFSee Noethe r note 1 1 The me asu reme nt of is d e s cr i be d in Append i x B

- 1 9 shy

FCON S 8 and CAPS 5 are pos it i ve wh i l e tho s e on MDS Pt - 1 and e are

negat i ve

Column s 1 and 2 of tab l e 1 pre s e nt the re s u lts from a 2SLS

regre s s i on of the sy stem conta i n i ng the re duc e d form income

equat i on ( 5 ) and th e FMG supp ly cu rve ( 4 ) In the MDINC equat i on

a l l the fore i gn and dome st i c re str i ct i ons show the pre d i ct e d

s i gns t h e FMG ent ry re st r i ct i ons inc r e a s e phy s i c i a n income

wh i le subs i d i es to us me d i c a l schoo l s de c r e a s e it The

p re d i ct i on of phy s i c i a n i ncome de r i ved here is used as the MDINC

va r i a b l e i n the FMG equ at i on s reported here in co lumn 2 It has

the expe cted pos it i ve ef f e ct A l l of the re str i ct i on va r i a b l e s

have the ant i c i p at e d negat i ve e f f e ct on FMG ent ry 2 4

Wh i l e th e re l a xat i on o f l i ce n s i ng and imm i grat i on laws

app arent ly d i d sh i ft the FMG supp l y cu rve to the r i g ht the

coe f f i c i e nt on the D7 6 va r i a b l e a l s o su ggests th at the cu rve has

s h i fted ba c k in rec e nt ye a rs In f a ct the coe f f i c i e nt on D7 6 i s

a lmost a s large a s that on D7 1 F rom th e est i mated FMG equ at i on s

and the pre d i ct i ons o f MDINC we c a n tra ce th e moveme nt of the

F MG suppl y cu rve ove r time F i gure 5 tra c e s its moveme nt It

s hows that afte r rema i n i ng re l at i ve ly stat i o n a ry unt i l th e lat e

1 9 60 s the supp l y sh i fted out subst a nt i a l ly i n the f i rst ha l f of

the 1 9 70 s The outw a rd moveme nt cont i nu e d unt i l about 1 9 7 5 du e

2 4 E ither i n c l u d i ng TNGTM i n the i ncome equ at i on or not i nc l u d i ng it in the FMG equat i on doe s not ch ange the re su lt s s i g n i f i c a nt ly The s imp le co rre lat i on betwe e n TOC and TNGTM e xc e e d s 9

- 20 shy

This re duced form income equation is de rive d from demand and

c o s t equa tions the unde r lying mode l for which is des c ribe d in

g re a te r de t ai l in Appe ndi x A I t uses pe r capita rea l in come

( INCM ) and the pe rce n t of the popu l a tion co ve re d by pu b lic insu r shy

a nce (PUB I NS ) as dem a n d va riable s 2 1 They shou l d both have a

positive e f fect on phy sician income Margina l cos t is me a s u red

as the tot a l opp ortunity cos t (TOC ) of be coming a physicia n

based on the as s ump tion tha t the ne xt be s t a l te rna tive is workin g

f o l lowin g fou r yea r s of co l le ge a n d ad j u s ted f o r the time spe nt

2 2in me di c a l schoo l and re side n cy prog r ams It shou l d a l s o be

p ositive l y co rre l a te d wit h phy sician income V aria b le s tha t

mea s u re fede ra l subsidies gra nted t o us medica l schoo l s bo th

f or con s t ru ction (CON S 8 ) and as capitation grants (CAPS 5 ) a re

a ls o in corporated and wil l be dis c u s se d fu rthe r in Section I I I

be low Whi le a true re s t ric tion on a dome s tic f ringe shou ld

r aise phy sician income since these are subsidie s they ease

e n t ry of the dome s tic f ringe and he nce lowe r phy sician in come

Fina l ly phy sician income is a f fe c te d by di f fe re n ce s be twee n

a c tu a l ent ry int o the profession and th at pre dicted by the

dominant firm As a re lated pape r shows this error ca n be

2 l O the r demand va ri a b le s we re a l s o used in Noe t he r sup r a notes 1 and 1 1 and these appe a re d mos t signific a n t Hence othe r variable s su ch as edu ca tion and age we re omit ted to avoid mu lti-co l line a rity prob lems with other independe n t va ri a b le s

2 2 The de riva tion and me a s u reme nt of TOC are fu l ly de s c ribe d in N oe t he r supra note 1 1 D at a sou rce s for TOC and the other v a ria bles used are lis ted in Appendi x B

- 1 8shy

supra

re duced to incorrect pre d i ct i ons about FMG ent ry wh i ch we de note

e F 2 3 Thu s the est i mated equ at i on i s as fol lows

( 5 ) MDINCt = oo + 0 1 TOCt + 0 2 1 INCMt + 0 2 2 PUBINSt + 0 3 1 D7 1t

+ 03 2 D7 6t + 0 3 3 NOLICt + 0 3 4 CITI Zt + 0 4 1 CON S 8t

F 2 + 04 2 CAPS 5t + 0 5 MDSPt - 1 + 0 6 et + Et

Add i t i ona l ly the re du c e d form equat i on de s cr i b i ng FMG

e nt ry de r i ved by su bst itut i ng ( 5 ) i nto ( 4 ) is a l s o est imated

a s

( 6 ) F MGPt = + TOCt + 2 1 INCMt + 2 2 PUBINSt + 3 1 D7 1t0 1

+ 3 2 D7 6t + 3 3 NOLICt + 3 4 CITI Zt + 3 5 TNGTMt

F 3 + 4 1 CON S 8t + 4 2 CAPS 5t + 5 MDSPt - 1 + 6 et + Et

The ef f e cts of the FMG re str i ct i on s and opp o rtu n i ty cost on the

redu c e d form of FMG are i ndete rm i nate The i r d i r e ct ef f e ct is

obv i ou s ly ne gat i ve but th ey have a pos it i ve i n f l u e n c e on i ncome

w h i ch a l s o a f f e cts the dep e nd e nt va r i a b l e pos it i ve l y P re s umshy

a b l y the forme r ef f e ct is strong e r so that the net impact is

negat i ve The rema i n i ng va r i a b l e s have th e same i nf luence as

they do on MDINC i e th e coe f f i c i e nts on TOC INCM PUBINS

2 3 eFSee Noethe r note 1 1 The me asu reme nt of is d e s cr i be d in Append i x B

- 1 9 shy

FCON S 8 and CAPS 5 are pos it i ve wh i l e tho s e on MDS Pt - 1 and e are

negat i ve

Column s 1 and 2 of tab l e 1 pre s e nt the re s u lts from a 2SLS

regre s s i on of the sy stem conta i n i ng the re duc e d form income

equat i on ( 5 ) and th e FMG supp ly cu rve ( 4 ) In the MDINC equat i on

a l l the fore i gn and dome st i c re str i ct i ons show the pre d i ct e d

s i gns t h e FMG ent ry re st r i ct i ons inc r e a s e phy s i c i a n income

wh i le subs i d i es to us me d i c a l schoo l s de c r e a s e it The

p re d i ct i on of phy s i c i a n i ncome de r i ved here is used as the MDINC

va r i a b l e i n the FMG equ at i on s reported here in co lumn 2 It has

the expe cted pos it i ve ef f e ct A l l of the re str i ct i on va r i a b l e s

have the ant i c i p at e d negat i ve e f f e ct on FMG ent ry 2 4

Wh i l e th e re l a xat i on o f l i ce n s i ng and imm i grat i on laws

app arent ly d i d sh i ft the FMG supp l y cu rve to the r i g ht the

coe f f i c i e nt on the D7 6 va r i a b l e a l s o su ggests th at the cu rve has

s h i fted ba c k in rec e nt ye a rs In f a ct the coe f f i c i e nt on D7 6 i s

a lmost a s large a s that on D7 1 F rom th e est i mated FMG equ at i on s

and the pre d i ct i ons o f MDINC we c a n tra ce th e moveme nt of the

F MG suppl y cu rve ove r time F i gure 5 tra c e s its moveme nt It

s hows that afte r rema i n i ng re l at i ve ly stat i o n a ry unt i l th e lat e

1 9 60 s the supp l y sh i fted out subst a nt i a l ly i n the f i rst ha l f of

the 1 9 70 s The outw a rd moveme nt cont i nu e d unt i l about 1 9 7 5 du e

2 4 E ither i n c l u d i ng TNGTM i n the i ncome equ at i on or not i nc l u d i ng it in the FMG equat i on doe s not ch ange the re su lt s s i g n i f i c a nt ly The s imp le co rre lat i on betwe e n TOC and TNGTM e xc e e d s 9

- 20 shy

supra

re duced to incorrect pre d i ct i ons about FMG ent ry wh i ch we de note

e F 2 3 Thu s the est i mated equ at i on i s as fol lows

( 5 ) MDINCt = oo + 0 1 TOCt + 0 2 1 INCMt + 0 2 2 PUBINSt + 0 3 1 D7 1t

+ 03 2 D7 6t + 0 3 3 NOLICt + 0 3 4 CITI Zt + 0 4 1 CON S 8t

F 2 + 04 2 CAPS 5t + 0 5 MDSPt - 1 + 0 6 et + Et

Add i t i ona l ly the re du c e d form equat i on de s cr i b i ng FMG

e nt ry de r i ved by su bst itut i ng ( 5 ) i nto ( 4 ) is a l s o est imated

a s

( 6 ) F MGPt = + TOCt + 2 1 INCMt + 2 2 PUBINSt + 3 1 D7 1t0 1

+ 3 2 D7 6t + 3 3 NOLICt + 3 4 CITI Zt + 3 5 TNGTMt

F 3 + 4 1 CON S 8t + 4 2 CAPS 5t + 5 MDSPt - 1 + 6 et + Et

The ef f e cts of the FMG re str i ct i on s and opp o rtu n i ty cost on the

redu c e d form of FMG are i ndete rm i nate The i r d i r e ct ef f e ct is

obv i ou s ly ne gat i ve but th ey have a pos it i ve i n f l u e n c e on i ncome

w h i ch a l s o a f f e cts the dep e nd e nt va r i a b l e pos it i ve l y P re s umshy

a b l y the forme r ef f e ct is strong e r so that the net impact is

negat i ve The rema i n i ng va r i a b l e s have th e same i nf luence as

they do on MDINC i e th e coe f f i c i e nts on TOC INCM PUBINS

2 3 eFSee Noethe r note 1 1 The me asu reme nt of is d e s cr i be d in Append i x B

- 1 9 shy

FCON S 8 and CAPS 5 are pos it i ve wh i l e tho s e on MDS Pt - 1 and e are

negat i ve

Column s 1 and 2 of tab l e 1 pre s e nt the re s u lts from a 2SLS

regre s s i on of the sy stem conta i n i ng the re duc e d form income

equat i on ( 5 ) and th e FMG supp ly cu rve ( 4 ) In the MDINC equat i on

a l l the fore i gn and dome st i c re str i ct i ons show the pre d i ct e d

s i gns t h e FMG ent ry re st r i ct i ons inc r e a s e phy s i c i a n income

wh i le subs i d i es to us me d i c a l schoo l s de c r e a s e it The

p re d i ct i on of phy s i c i a n i ncome de r i ved here is used as the MDINC

va r i a b l e i n the FMG equ at i on s reported here in co lumn 2 It has

the expe cted pos it i ve ef f e ct A l l of the re str i ct i on va r i a b l e s

have the ant i c i p at e d negat i ve e f f e ct on FMG ent ry 2 4

Wh i l e th e re l a xat i on o f l i ce n s i ng and imm i grat i on laws

app arent ly d i d sh i ft the FMG supp l y cu rve to the r i g ht the

coe f f i c i e nt on the D7 6 va r i a b l e a l s o su ggests th at the cu rve has

s h i fted ba c k in rec e nt ye a rs In f a ct the coe f f i c i e nt on D7 6 i s

a lmost a s large a s that on D7 1 F rom th e est i mated FMG equ at i on s

and the pre d i ct i ons o f MDINC we c a n tra ce th e moveme nt of the

F MG suppl y cu rve ove r time F i gure 5 tra c e s its moveme nt It

s hows that afte r rema i n i ng re l at i ve ly stat i o n a ry unt i l th e lat e

1 9 60 s the supp l y sh i fted out subst a nt i a l ly i n the f i rst ha l f of

the 1 9 70 s The outw a rd moveme nt cont i nu e d unt i l about 1 9 7 5 du e

2 4 E ither i n c l u d i ng TNGTM i n the i ncome equ at i on or not i nc l u d i ng it in the FMG equat i on doe s not ch ange the re su lt s s i g n i f i c a nt ly The s imp le co rre lat i on betwe e n TOC and TNGTM e xc e e d s 9

- 20 shy

FCON S 8 and CAPS 5 are pos it i ve wh i l e tho s e on MDS Pt - 1 and e are

negat i ve

Column s 1 and 2 of tab l e 1 pre s e nt the re s u lts from a 2SLS

regre s s i on of the sy stem conta i n i ng the re duc e d form income

equat i on ( 5 ) and th e FMG supp ly cu rve ( 4 ) In the MDINC equat i on

a l l the fore i gn and dome st i c re str i ct i ons show the pre d i ct e d

s i gns t h e FMG ent ry re st r i ct i ons inc r e a s e phy s i c i a n income

wh i le subs i d i es to us me d i c a l schoo l s de c r e a s e it The

p re d i ct i on of phy s i c i a n i ncome de r i ved here is used as the MDINC

va r i a b l e i n the FMG equ at i on s reported here in co lumn 2 It has

the expe cted pos it i ve ef f e ct A l l of the re str i ct i on va r i a b l e s

have the ant i c i p at e d negat i ve e f f e ct on FMG ent ry 2 4

Wh i l e th e re l a xat i on o f l i ce n s i ng and imm i grat i on laws

app arent ly d i d sh i ft the FMG supp l y cu rve to the r i g ht the

coe f f i c i e nt on the D7 6 va r i a b l e a l s o su ggests th at the cu rve has

s h i fted ba c k in rec e nt ye a rs In f a ct the coe f f i c i e nt on D7 6 i s

a lmost a s large a s that on D7 1 F rom th e est i mated FMG equ at i on s

and the pre d i ct i ons o f MDINC we c a n tra ce th e moveme nt of the

F MG suppl y cu rve ove r time F i gure 5 tra c e s its moveme nt It

s hows that afte r rema i n i ng re l at i ve ly stat i o n a ry unt i l th e lat e

1 9 60 s the supp l y sh i fted out subst a nt i a l ly i n the f i rst ha l f of

the 1 9 70 s The outw a rd moveme nt cont i nu e d unt i l about 1 9 7 5 du e

2 4 E ither i n c l u d i ng TNGTM i n the i ncome equ at i on or not i nc l u d i ng it in the FMG equat i on doe s not ch ange the re su lt s s i g n i f i c a nt ly The s imp le co rre lat i on betwe e n TOC and TNGTM e xc e e d s 9

- 20 shy

middot=--middot-=--=-=

- 2 1 -

TABLE l

FMG SU PPLY CURVE EQUATION ( COLUMNS

STRUCTU RAL ( 2 SLS ) ESTIMATION WITH INCOME 1amp 2 ) REDUCED FORM ESTIMATION ( COLUMN 3 )

1 9 4 6 - 1 9 8 1

Dependent Va r i a ble MD INC FMGP FMGP

( l ) ( 2 ) ( 3 )

INTERCEPT - 1 8 0 6 3 2 2 4 6 ( - l ) 6 2 7 5 ( - l ) ( - 1 3 7 ) ( 2 9 6 ) ( l 6 8 )

TOC 1 0 3 9 1 7 4 4 ( - 4 ) ( bull 0 2 )( 4 9 7 )

4 9 1 3 ( - 2 ) INCM 6 9 50 ( 4 6 )

4 8 9 2 ( - 4 )

( 2 4 1 )

PUB INS 1 6 1 8 ( l 7 )

- 2 6 8 7 ( - l )

( 2 1 3 )

MDSP ( t - 1 ) - 1 2 2 8 9 ( - 1 2 ) ( - 9 9 )

e 2 2 4 9 3 4 F - 5 1 1 3 ( - l ) ( - 1 2 4 ) ( 2 20 )

- 5 30 4 ( - 2 ) ( - 3 5 3 )

CON S 8 - 1 7 9 1 8 5 ( - 2 9 )

CAPS 5 - 1 5 6 5 4 5 ( - 6 2 )

D7 l 4 3 9 6 - 1 3 5 9 ( - l ) - 1 4 6 7 ( - l ) ( bull 4 2 ) ( - 4 4 9 ) ( - 4 6 6 )

D 7 6 4 7 5 3 1 - 100 3 ( - l - 10 3 7 ( - 1 ) ( 2 1 2 ) ( - 4 6 4 ) ( - 1 5 8 )

NOLIC 1 2 6 9 - 3 6 0 3 ( - 3 ) - 3 6 7 6 ( - 3 ) ( 1 9 1 ) ( - 1 8 6 ) ( - 1 4 6 )

CITI Z 6 7 6 8 ( - 1 ) - 10 2 9 ( - 3 ) - 2 2 9 5 ( - 3 ) ( 1 6 2 ) ( - 9 1 ) ( - 1 4 3 )

TAB LE 1--Cont i nued

M D INC ( 1 )

Dependent Var i a b le FMGP ( 2 )

FMGP ( 3 )

TNGTM - 7 9 9 7 ( -2 ) ( - 1 4 7 )

- 7 1 4 8 ( - 1 ) ( - 1 00 )

MD INC 4 1 2 8 ( - 3 ) ( 1 6 5 )

S SE 1 2 5 3 2 3 9 5 ( - 3 ) 1 6 8 6 ( - 3 )

R 2 9 9 2 9 9 2 8 3 9 4 2 8

D W 1 9 2 1 6 2

p - 0 6

Notes t-stat ist i cs i n parentheses ( -n ) imp l i es that 1o nreported coe f f i ce nt is t imes act u a l ma gn itude MDINC in

FMGP re gress i on ( 2 ) is pre d i ct i on f rom MDINC re gress i ons ( 1 ) p = co rre lat ion coe f f i c i e nt used in Pra is -W i nste n est imat i on

- 2 2 shy

FIGURE 5

E st i mat e d FMG Supp ly Cu rve s for Se l e ct e d Yea rs ( f rom 2SLS Reg r e s s i ons w it h yea r s ma rk ed at top of cu r ve s )

$ MD

40000

30000

20000

10000

1000 2000 3000 4000 5000 6000 F 1G

- 2 3 shy

to ch an g i ng l i ce ns i ng laws and imp l i e s an i nc r e a s e of 4000 FMGs

at any g i ve n i ncome le ve l betw e e n 1 9 6 8 and 19 7 5 F i na l l y when

the imm i grat i o n law was ti ghtened i n 1 9 7 7 th e supp l y cu rve

s h i fted ba c k but not a l l th e way to its or i g i na l pos it i on The

1 9 8 1 FMG supp ly cu rve l i e s about hal fway betwe e n its late 1 9 60s

a n d ea r ly - 1 9 70s pos it i on s 2 5

Re s u lt s f rom est imat i n g the unre str i cte d redu c e d form

equat ion ( 6 ) a re pre s ented in co lumn 3 of Tab l e 1 Med i c a l

s choo l constru ct i o n fund i ng ( CONS 8 ) i s used to rep r e s e nt the

dome st i c f r i nge su bs i dy CAPS 5 was al s o used i n some re gre s s i on s

but its h i gh co rre l at i on w ith CONS 8 ( 9 1 ) a s we l l a s w ith th e

F MG re str i ct i ons created prob lems in i d e nt i f y i ng coe f f i c i e nt

e st imates The re g re s s ion was est imated us i ng a Pra i s -W i nsten

tra ns f ormat i on to cor r e ct autocorre l at i on Wh i le it is

i mpos s i b l e to det e rm i ne the net e f f e ct s of th e re str i ct i on

var i a b l e s f rom the theoret i c a l mod e l be c a u s e of the of f s ett i ng

i mp acts of the i r i n f luences on phy s i c i a n i ncome the emp i r i c a l

r e s u lts show at le a st the i r s i g ns ( a nd often the i r coe f f i cient s )

to be s im i l a r to the 2SLS re su lts

F i na l ly we can comp a re what ou r va r i ou s est i mates ind i cate

h as bee n the cumu l at i ve e f f e ct of ch anges in the l i ce n s i ng and

25 The t ime path der i ve d us i ng the unreport e d D6 8 reg re s s i on is s im i l a r exc ept that the outwa rd exp ans i o n be g i n s ea r l i e r and i s mor e gradua l

- 2 4 shy

imm i g r at i on law s s i nc e 1 9 6 5 We me a s u re the ef f e ct of each va r i shy

a b l e i n yea r t by tak i ng the d i f f e rence betwe e n its actu a l ma gn i shy

tude in yea r t and th e va l u e it had in 1 9 6 5 t i me s its est i mated

coe f f i c i e nt Table 2 shows th e cumu lat i v e ef f e ct from 19 6 6

through 1 9 8 1 of th e l i c ens i n g and imm i grat i on reg u l at i ons

e f f e ct s for each of ou r re gres s i ons The re su lt s su ggest that

re l a xat i o n of the s e law s ena b l e d somewh e re f r om 3 3 000 to 4 5 000

add it i ona l FMG s to obta i n us me d i c a l l i c e n s e s du r i ng th e

f i fteen yea r pe r i od The equat i on s suggest a les s e r e f f e ct f rom

the l i c e n s i ng va r i a b l e s equa l i ng 12 000 - 2 2 000 and a net pos ishy

t i ve impact ( rememb e r that the 1 9 7 6 law re str i cted ent ry aga i n )

f rom imm i g r at i on po l i cy of 2 1 000 - 2 4 000 The D7 l imm i g r at i on

va r i ab le a l l owed about 3 000 entra nt s annu a l ly wh i l e D7 6 w ithhe l d

about 2 000 O f the l i ce n s i ng va r i ab le s CITIZ is most i n f l u e nshy

t i a l it ac cou nt s for 60 to 7 5 per c e nt of the tota l l i ce n s i ng

e f f ect The tota l ef f ect s i nc e 1 9 6 5 of loo s e n e d reg u l at i o n s on

FMG ent ry ha s bee n substant i a l equa l i ng 4 5 to 6 5 pe r c e nt of the

app rox i mate l y 70 000 FMGs th at have be come l i c e n s e d i n the us

s i nce 1 9 6 5

P roba b l y the pr i ma ry imp etu s for re l a xat i on of the

Imm i g r a nt and V i s a law s came f rom the pe rce i ve d shortage of

p hy s i c i a ns in the us in the 1 9 60s Moreove r the large

e xp a ns i on of hos p ita l i nte rn s h ip and re s i de ncy tra i n i ng pos it ions

f o l low i ng Wor l d Wa r II pa rt ly du e to the H i l l -Bu rton hosp ita l

constru ct i on prog ram created a dema n d for gra du ate me d i ca l

student s wh i ch th e us schoo l s grad u ates d i d not f i l l The AMA

- 2 5 shy

TAB LE 2

IMPACT OF RELAXATION OF FMG ENTRY RE STRICTIONS FROM 1 9 6 5 LEVEL ON NUMB E R OF NEW FMG LICENSE E S

1 9 6 6 - 1 9 8 1 CUMU LATIVE EFF ECT

2 SLS Redu ced

Form

L i ce ns i ng

Imm i g r at i on

Tota l

1 3 2 7 3

2 1 3 10

3 4 5 8 3

2 1 9 40

2 3 5 1 6

4 5 4 5 6

- 2 6 shy

has certa i n ly advo cated exc lus i on a ry po l i c i es It opposed

r e l a xat i on of the imm i grat i on law in the 1 9 60 s and ba ck in

1 9 3 8 th e AMA House of De l e g ates res o l ve d th at us cit ize ns h ip

be i n c lu ded as a prerequ is ite i n all state me d i ca l l i cens i ng

statutes 2 6 The t i ghten i ng of the imm i grat i on law i n 1 9 7 6 came

at a time when U S schoo ls we re ge nerat i ng a substant i a l ly

l a rge r number of annu a l gra du ates th an they ha d be e n at the

beg i nn i ng of the de cade More over re i mpos it i on of str i ct state

l i c e ns i ng laws has not oc curre d Such act i on is perhaps too

cost ly for th e AMA requ i r i ng 50 sep a rate lobby i ng ef forts

The de c l i ne i n FMG ent ry in the last few yea rs does not

i mp ly that compet it i o n i n the med i ca l prof ess i on has abated

u s me d i ca l scho o ls dou b l e d the i r annu a l output betwe e n 1 9 6 5 and

1 9 80 and in the last few yea rs have mo re tha n ma de up for th e

reduced numb e r of FMG s obta i n i ng us l i c e nses In the ne xt

sect i on we w i l l exam i ne th e det e rm i nants of us new l i ce nsees

i n ord e r to iso lat e th e domest i c compet it i ve f r i nge

III DOMESTIC COMP ETITIVE FRINGE ENTRY

M easur i ng the second comp et it i ve f r i nge th e add it i ona l us

g raduates resu lt i ng so le ly f rom fede ra l ly su bs i d ized me d i c a l

school growth is more comp l i c ated To att r i bute al l o f the 100

p e r c e nt growth betwe e n 1 9 6 5 and 19 80 in the annu a l numb e r of

2 6 I re n e Butte r and Rebe c c a G Sweet L i c e nsu re of Fore i gn M ed i c a l Graduates An H istor i ca l Perspe ct i ve 5 5 Mi l b a nk M emor i a l Fund Qua rt e r ly 3 1 5 ( 1 9 7 7 )

- 2 7 shy

graduates to an ent ry ba rr i e r bre a k down is su re ly an ove rstate shy

ment The adve nt of the Me d i c a i d and Med i c a r e progr ams in 1 9 6 5

i nc r e a s e d dema nd as d i d h i g h e r consume r i ncome s unt i l th e late

1 9 70s There f ore us i ng the dom i nant f i rm mode l pa rt of the

i nc r e a s e in me d i ca l s choo l output ca n proba b l y be att r i buted to

the opt i ma l ad j u stme nt of a ca rte l to growth in dema nd2 7 S i nce

d emand cond it i ons changed after 19 6 5 we ca nnot s imp ly pro j e ct

g rowth in the dom i na nt f i rm from its pre - 1 9 6 5 be h a v i or

If we pos it the ma rg i na l cost curve for phy s i c i a ns as hor i shy

zont a l at the le ve l of tot a l opp ortu n ity co st up to a constra i nt

2 8d et e rm i ne d by th e numbe r of me d i c a l schoo l sp a c e s and ve rt i ca l

the r e a ft e r a growth i n th e dome st i c compet it i ve f r i nge imp l i e s a

r i ghtw ard sh i f t of th e schoo l cap a c i ty constra i nt That is in

t h i s cas e ne ither a lowe r i ng of the opportu n ity cost nor a ri s e

i n phy s i c i a n i ncome ca n le ad to incr e a s e d US gr aduate ent ry

u n l e s s the med i ca l s choo l cap a c ity is increas e d Conve rs e l y th e

a ctu a l re l a xat i on of th e entry ba r r i e r doe s not le a d to a reduced

opp ortu n ity cost for thos e ac c epted i nto me d i c a l schoo l ie

the hor izont a l segme nt of the cu rve is not lowe red

2 7 It is not c l e a r that an ex i st i ng numbe r of ( p e r f ect l y c a rte l ized ) phy s i c i a ns wou l d want to permit entry o f new p hy s i c i a ns rather th an exp a n d i ng the i r ow n output Such a de s i re c a n be rat i on a l ized as an attempt to pre ve nt co nsume r s from a ct i ve l y seek i ng subst itutes

2 8 G i ve n the sma l l proport i on of co llege gr a du ates that each y e a rs med i c a l schoo l c l a s s rep r e s e nts th i s seems a re a s on a b l e a s s umpt i on

- 2 8 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

TAB LE 1--Cont i nued

M D INC ( 1 )

Dependent Var i a b le FMGP ( 2 )

FMGP ( 3 )

TNGTM - 7 9 9 7 ( -2 ) ( - 1 4 7 )

- 7 1 4 8 ( - 1 ) ( - 1 00 )

MD INC 4 1 2 8 ( - 3 ) ( 1 6 5 )

S SE 1 2 5 3 2 3 9 5 ( - 3 ) 1 6 8 6 ( - 3 )

R 2 9 9 2 9 9 2 8 3 9 4 2 8

D W 1 9 2 1 6 2

p - 0 6

Notes t-stat ist i cs i n parentheses ( -n ) imp l i es that 1o nreported coe f f i ce nt is t imes act u a l ma gn itude MDINC in

FMGP re gress i on ( 2 ) is pre d i ct i on f rom MDINC re gress i ons ( 1 ) p = co rre lat ion coe f f i c i e nt used in Pra is -W i nste n est imat i on

- 2 2 shy

FIGURE 5

E st i mat e d FMG Supp ly Cu rve s for Se l e ct e d Yea rs ( f rom 2SLS Reg r e s s i ons w it h yea r s ma rk ed at top of cu r ve s )

$ MD

40000

30000

20000

10000

1000 2000 3000 4000 5000 6000 F 1G

- 2 3 shy

to ch an g i ng l i ce ns i ng laws and imp l i e s an i nc r e a s e of 4000 FMGs

at any g i ve n i ncome le ve l betw e e n 1 9 6 8 and 19 7 5 F i na l l y when

the imm i grat i o n law was ti ghtened i n 1 9 7 7 th e supp l y cu rve

s h i fted ba c k but not a l l th e way to its or i g i na l pos it i on The

1 9 8 1 FMG supp ly cu rve l i e s about hal fway betwe e n its late 1 9 60s

a n d ea r ly - 1 9 70s pos it i on s 2 5

Re s u lt s f rom est imat i n g the unre str i cte d redu c e d form

equat ion ( 6 ) a re pre s ented in co lumn 3 of Tab l e 1 Med i c a l

s choo l constru ct i o n fund i ng ( CONS 8 ) i s used to rep r e s e nt the

dome st i c f r i nge su bs i dy CAPS 5 was al s o used i n some re gre s s i on s

but its h i gh co rre l at i on w ith CONS 8 ( 9 1 ) a s we l l a s w ith th e

F MG re str i ct i ons created prob lems in i d e nt i f y i ng coe f f i c i e nt

e st imates The re g re s s ion was est imated us i ng a Pra i s -W i nsten

tra ns f ormat i on to cor r e ct autocorre l at i on Wh i le it is

i mpos s i b l e to det e rm i ne the net e f f e ct s of th e re str i ct i on

var i a b l e s f rom the theoret i c a l mod e l be c a u s e of the of f s ett i ng

i mp acts of the i r i n f luences on phy s i c i a n i ncome the emp i r i c a l

r e s u lts show at le a st the i r s i g ns ( a nd often the i r coe f f i cient s )

to be s im i l a r to the 2SLS re su lts

F i na l ly we can comp a re what ou r va r i ou s est i mates ind i cate

h as bee n the cumu l at i ve e f f e ct of ch anges in the l i ce n s i ng and

25 The t ime path der i ve d us i ng the unreport e d D6 8 reg re s s i on is s im i l a r exc ept that the outwa rd exp ans i o n be g i n s ea r l i e r and i s mor e gradua l

- 2 4 shy

imm i g r at i on law s s i nc e 1 9 6 5 We me a s u re the ef f e ct of each va r i shy

a b l e i n yea r t by tak i ng the d i f f e rence betwe e n its actu a l ma gn i shy

tude in yea r t and th e va l u e it had in 1 9 6 5 t i me s its est i mated

coe f f i c i e nt Table 2 shows th e cumu lat i v e ef f e ct from 19 6 6

through 1 9 8 1 of th e l i c ens i n g and imm i grat i on reg u l at i ons

e f f e ct s for each of ou r re gres s i ons The re su lt s su ggest that

re l a xat i o n of the s e law s ena b l e d somewh e re f r om 3 3 000 to 4 5 000

add it i ona l FMG s to obta i n us me d i c a l l i c e n s e s du r i ng th e

f i fteen yea r pe r i od The equat i on s suggest a les s e r e f f e ct f rom

the l i c e n s i ng va r i a b l e s equa l i ng 12 000 - 2 2 000 and a net pos ishy

t i ve impact ( rememb e r that the 1 9 7 6 law re str i cted ent ry aga i n )

f rom imm i g r at i on po l i cy of 2 1 000 - 2 4 000 The D7 l imm i g r at i on

va r i ab le a l l owed about 3 000 entra nt s annu a l ly wh i l e D7 6 w ithhe l d

about 2 000 O f the l i ce n s i ng va r i ab le s CITIZ is most i n f l u e nshy

t i a l it ac cou nt s for 60 to 7 5 per c e nt of the tota l l i ce n s i ng

e f f ect The tota l ef f ect s i nc e 1 9 6 5 of loo s e n e d reg u l at i o n s on

FMG ent ry ha s bee n substant i a l equa l i ng 4 5 to 6 5 pe r c e nt of the

app rox i mate l y 70 000 FMGs th at have be come l i c e n s e d i n the us

s i nce 1 9 6 5

P roba b l y the pr i ma ry imp etu s for re l a xat i on of the

Imm i g r a nt and V i s a law s came f rom the pe rce i ve d shortage of

p hy s i c i a ns in the us in the 1 9 60s Moreove r the large

e xp a ns i on of hos p ita l i nte rn s h ip and re s i de ncy tra i n i ng pos it ions

f o l low i ng Wor l d Wa r II pa rt ly du e to the H i l l -Bu rton hosp ita l

constru ct i on prog ram created a dema n d for gra du ate me d i ca l

student s wh i ch th e us schoo l s grad u ates d i d not f i l l The AMA

- 2 5 shy

TAB LE 2

IMPACT OF RELAXATION OF FMG ENTRY RE STRICTIONS FROM 1 9 6 5 LEVEL ON NUMB E R OF NEW FMG LICENSE E S

1 9 6 6 - 1 9 8 1 CUMU LATIVE EFF ECT

2 SLS Redu ced

Form

L i ce ns i ng

Imm i g r at i on

Tota l

1 3 2 7 3

2 1 3 10

3 4 5 8 3

2 1 9 40

2 3 5 1 6

4 5 4 5 6

- 2 6 shy

has certa i n ly advo cated exc lus i on a ry po l i c i es It opposed

r e l a xat i on of the imm i grat i on law in the 1 9 60 s and ba ck in

1 9 3 8 th e AMA House of De l e g ates res o l ve d th at us cit ize ns h ip

be i n c lu ded as a prerequ is ite i n all state me d i ca l l i cens i ng

statutes 2 6 The t i ghten i ng of the imm i grat i on law i n 1 9 7 6 came

at a time when U S schoo ls we re ge nerat i ng a substant i a l ly

l a rge r number of annu a l gra du ates th an they ha d be e n at the

beg i nn i ng of the de cade More over re i mpos it i on of str i ct state

l i c e ns i ng laws has not oc curre d Such act i on is perhaps too

cost ly for th e AMA requ i r i ng 50 sep a rate lobby i ng ef forts

The de c l i ne i n FMG ent ry in the last few yea rs does not

i mp ly that compet it i o n i n the med i ca l prof ess i on has abated

u s me d i ca l scho o ls dou b l e d the i r annu a l output betwe e n 1 9 6 5 and

1 9 80 and in the last few yea rs have mo re tha n ma de up for th e

reduced numb e r of FMG s obta i n i ng us l i c e nses In the ne xt

sect i on we w i l l exam i ne th e det e rm i nants of us new l i ce nsees

i n ord e r to iso lat e th e domest i c compet it i ve f r i nge

III DOMESTIC COMP ETITIVE FRINGE ENTRY

M easur i ng the second comp et it i ve f r i nge th e add it i ona l us

g raduates resu lt i ng so le ly f rom fede ra l ly su bs i d ized me d i c a l

school growth is more comp l i c ated To att r i bute al l o f the 100

p e r c e nt growth betwe e n 1 9 6 5 and 19 80 in the annu a l numb e r of

2 6 I re n e Butte r and Rebe c c a G Sweet L i c e nsu re of Fore i gn M ed i c a l Graduates An H istor i ca l Perspe ct i ve 5 5 Mi l b a nk M emor i a l Fund Qua rt e r ly 3 1 5 ( 1 9 7 7 )

- 2 7 shy

graduates to an ent ry ba rr i e r bre a k down is su re ly an ove rstate shy

ment The adve nt of the Me d i c a i d and Med i c a r e progr ams in 1 9 6 5

i nc r e a s e d dema nd as d i d h i g h e r consume r i ncome s unt i l th e late

1 9 70s There f ore us i ng the dom i nant f i rm mode l pa rt of the

i nc r e a s e in me d i ca l s choo l output ca n proba b l y be att r i buted to

the opt i ma l ad j u stme nt of a ca rte l to growth in dema nd2 7 S i nce

d emand cond it i ons changed after 19 6 5 we ca nnot s imp ly pro j e ct

g rowth in the dom i na nt f i rm from its pre - 1 9 6 5 be h a v i or

If we pos it the ma rg i na l cost curve for phy s i c i a ns as hor i shy

zont a l at the le ve l of tot a l opp ortu n ity co st up to a constra i nt

2 8d et e rm i ne d by th e numbe r of me d i c a l schoo l sp a c e s and ve rt i ca l

the r e a ft e r a growth i n th e dome st i c compet it i ve f r i nge imp l i e s a

r i ghtw ard sh i f t of th e schoo l cap a c i ty constra i nt That is in

t h i s cas e ne ither a lowe r i ng of the opportu n ity cost nor a ri s e

i n phy s i c i a n i ncome ca n le ad to incr e a s e d US gr aduate ent ry

u n l e s s the med i ca l s choo l cap a c ity is increas e d Conve rs e l y th e

a ctu a l re l a xat i on of th e entry ba r r i e r doe s not le a d to a reduced

opp ortu n ity cost for thos e ac c epted i nto me d i c a l schoo l ie

the hor izont a l segme nt of the cu rve is not lowe red

2 7 It is not c l e a r that an ex i st i ng numbe r of ( p e r f ect l y c a rte l ized ) phy s i c i a ns wou l d want to permit entry o f new p hy s i c i a ns rather th an exp a n d i ng the i r ow n output Such a de s i re c a n be rat i on a l ized as an attempt to pre ve nt co nsume r s from a ct i ve l y seek i ng subst itutes

2 8 G i ve n the sma l l proport i on of co llege gr a du ates that each y e a rs med i c a l schoo l c l a s s rep r e s e nts th i s seems a re a s on a b l e a s s umpt i on

- 2 8 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

FIGURE 5

E st i mat e d FMG Supp ly Cu rve s for Se l e ct e d Yea rs ( f rom 2SLS Reg r e s s i ons w it h yea r s ma rk ed at top of cu r ve s )

$ MD

40000

30000

20000

10000

1000 2000 3000 4000 5000 6000 F 1G

- 2 3 shy

to ch an g i ng l i ce ns i ng laws and imp l i e s an i nc r e a s e of 4000 FMGs

at any g i ve n i ncome le ve l betw e e n 1 9 6 8 and 19 7 5 F i na l l y when

the imm i grat i o n law was ti ghtened i n 1 9 7 7 th e supp l y cu rve

s h i fted ba c k but not a l l th e way to its or i g i na l pos it i on The

1 9 8 1 FMG supp ly cu rve l i e s about hal fway betwe e n its late 1 9 60s

a n d ea r ly - 1 9 70s pos it i on s 2 5

Re s u lt s f rom est imat i n g the unre str i cte d redu c e d form

equat ion ( 6 ) a re pre s ented in co lumn 3 of Tab l e 1 Med i c a l

s choo l constru ct i o n fund i ng ( CONS 8 ) i s used to rep r e s e nt the

dome st i c f r i nge su bs i dy CAPS 5 was al s o used i n some re gre s s i on s

but its h i gh co rre l at i on w ith CONS 8 ( 9 1 ) a s we l l a s w ith th e

F MG re str i ct i ons created prob lems in i d e nt i f y i ng coe f f i c i e nt

e st imates The re g re s s ion was est imated us i ng a Pra i s -W i nsten

tra ns f ormat i on to cor r e ct autocorre l at i on Wh i le it is

i mpos s i b l e to det e rm i ne the net e f f e ct s of th e re str i ct i on

var i a b l e s f rom the theoret i c a l mod e l be c a u s e of the of f s ett i ng

i mp acts of the i r i n f luences on phy s i c i a n i ncome the emp i r i c a l

r e s u lts show at le a st the i r s i g ns ( a nd often the i r coe f f i cient s )

to be s im i l a r to the 2SLS re su lts

F i na l ly we can comp a re what ou r va r i ou s est i mates ind i cate

h as bee n the cumu l at i ve e f f e ct of ch anges in the l i ce n s i ng and

25 The t ime path der i ve d us i ng the unreport e d D6 8 reg re s s i on is s im i l a r exc ept that the outwa rd exp ans i o n be g i n s ea r l i e r and i s mor e gradua l

- 2 4 shy

imm i g r at i on law s s i nc e 1 9 6 5 We me a s u re the ef f e ct of each va r i shy

a b l e i n yea r t by tak i ng the d i f f e rence betwe e n its actu a l ma gn i shy

tude in yea r t and th e va l u e it had in 1 9 6 5 t i me s its est i mated

coe f f i c i e nt Table 2 shows th e cumu lat i v e ef f e ct from 19 6 6

through 1 9 8 1 of th e l i c ens i n g and imm i grat i on reg u l at i ons

e f f e ct s for each of ou r re gres s i ons The re su lt s su ggest that

re l a xat i o n of the s e law s ena b l e d somewh e re f r om 3 3 000 to 4 5 000

add it i ona l FMG s to obta i n us me d i c a l l i c e n s e s du r i ng th e

f i fteen yea r pe r i od The equat i on s suggest a les s e r e f f e ct f rom

the l i c e n s i ng va r i a b l e s equa l i ng 12 000 - 2 2 000 and a net pos ishy

t i ve impact ( rememb e r that the 1 9 7 6 law re str i cted ent ry aga i n )

f rom imm i g r at i on po l i cy of 2 1 000 - 2 4 000 The D7 l imm i g r at i on

va r i ab le a l l owed about 3 000 entra nt s annu a l ly wh i l e D7 6 w ithhe l d

about 2 000 O f the l i ce n s i ng va r i ab le s CITIZ is most i n f l u e nshy

t i a l it ac cou nt s for 60 to 7 5 per c e nt of the tota l l i ce n s i ng

e f f ect The tota l ef f ect s i nc e 1 9 6 5 of loo s e n e d reg u l at i o n s on

FMG ent ry ha s bee n substant i a l equa l i ng 4 5 to 6 5 pe r c e nt of the

app rox i mate l y 70 000 FMGs th at have be come l i c e n s e d i n the us

s i nce 1 9 6 5

P roba b l y the pr i ma ry imp etu s for re l a xat i on of the

Imm i g r a nt and V i s a law s came f rom the pe rce i ve d shortage of

p hy s i c i a ns in the us in the 1 9 60s Moreove r the large

e xp a ns i on of hos p ita l i nte rn s h ip and re s i de ncy tra i n i ng pos it ions

f o l low i ng Wor l d Wa r II pa rt ly du e to the H i l l -Bu rton hosp ita l

constru ct i on prog ram created a dema n d for gra du ate me d i ca l

student s wh i ch th e us schoo l s grad u ates d i d not f i l l The AMA

- 2 5 shy

TAB LE 2

IMPACT OF RELAXATION OF FMG ENTRY RE STRICTIONS FROM 1 9 6 5 LEVEL ON NUMB E R OF NEW FMG LICENSE E S

1 9 6 6 - 1 9 8 1 CUMU LATIVE EFF ECT

2 SLS Redu ced

Form

L i ce ns i ng

Imm i g r at i on

Tota l

1 3 2 7 3

2 1 3 10

3 4 5 8 3

2 1 9 40

2 3 5 1 6

4 5 4 5 6

- 2 6 shy

has certa i n ly advo cated exc lus i on a ry po l i c i es It opposed

r e l a xat i on of the imm i grat i on law in the 1 9 60 s and ba ck in

1 9 3 8 th e AMA House of De l e g ates res o l ve d th at us cit ize ns h ip

be i n c lu ded as a prerequ is ite i n all state me d i ca l l i cens i ng

statutes 2 6 The t i ghten i ng of the imm i grat i on law i n 1 9 7 6 came

at a time when U S schoo ls we re ge nerat i ng a substant i a l ly

l a rge r number of annu a l gra du ates th an they ha d be e n at the

beg i nn i ng of the de cade More over re i mpos it i on of str i ct state

l i c e ns i ng laws has not oc curre d Such act i on is perhaps too

cost ly for th e AMA requ i r i ng 50 sep a rate lobby i ng ef forts

The de c l i ne i n FMG ent ry in the last few yea rs does not

i mp ly that compet it i o n i n the med i ca l prof ess i on has abated

u s me d i ca l scho o ls dou b l e d the i r annu a l output betwe e n 1 9 6 5 and

1 9 80 and in the last few yea rs have mo re tha n ma de up for th e

reduced numb e r of FMG s obta i n i ng us l i c e nses In the ne xt

sect i on we w i l l exam i ne th e det e rm i nants of us new l i ce nsees

i n ord e r to iso lat e th e domest i c compet it i ve f r i nge

III DOMESTIC COMP ETITIVE FRINGE ENTRY

M easur i ng the second comp et it i ve f r i nge th e add it i ona l us

g raduates resu lt i ng so le ly f rom fede ra l ly su bs i d ized me d i c a l

school growth is more comp l i c ated To att r i bute al l o f the 100

p e r c e nt growth betwe e n 1 9 6 5 and 19 80 in the annu a l numb e r of

2 6 I re n e Butte r and Rebe c c a G Sweet L i c e nsu re of Fore i gn M ed i c a l Graduates An H istor i ca l Perspe ct i ve 5 5 Mi l b a nk M emor i a l Fund Qua rt e r ly 3 1 5 ( 1 9 7 7 )

- 2 7 shy

graduates to an ent ry ba rr i e r bre a k down is su re ly an ove rstate shy

ment The adve nt of the Me d i c a i d and Med i c a r e progr ams in 1 9 6 5

i nc r e a s e d dema nd as d i d h i g h e r consume r i ncome s unt i l th e late

1 9 70s There f ore us i ng the dom i nant f i rm mode l pa rt of the

i nc r e a s e in me d i ca l s choo l output ca n proba b l y be att r i buted to

the opt i ma l ad j u stme nt of a ca rte l to growth in dema nd2 7 S i nce

d emand cond it i ons changed after 19 6 5 we ca nnot s imp ly pro j e ct

g rowth in the dom i na nt f i rm from its pre - 1 9 6 5 be h a v i or

If we pos it the ma rg i na l cost curve for phy s i c i a ns as hor i shy

zont a l at the le ve l of tot a l opp ortu n ity co st up to a constra i nt

2 8d et e rm i ne d by th e numbe r of me d i c a l schoo l sp a c e s and ve rt i ca l

the r e a ft e r a growth i n th e dome st i c compet it i ve f r i nge imp l i e s a

r i ghtw ard sh i f t of th e schoo l cap a c i ty constra i nt That is in

t h i s cas e ne ither a lowe r i ng of the opportu n ity cost nor a ri s e

i n phy s i c i a n i ncome ca n le ad to incr e a s e d US gr aduate ent ry

u n l e s s the med i ca l s choo l cap a c ity is increas e d Conve rs e l y th e

a ctu a l re l a xat i on of th e entry ba r r i e r doe s not le a d to a reduced

opp ortu n ity cost for thos e ac c epted i nto me d i c a l schoo l ie

the hor izont a l segme nt of the cu rve is not lowe red

2 7 It is not c l e a r that an ex i st i ng numbe r of ( p e r f ect l y c a rte l ized ) phy s i c i a ns wou l d want to permit entry o f new p hy s i c i a ns rather th an exp a n d i ng the i r ow n output Such a de s i re c a n be rat i on a l ized as an attempt to pre ve nt co nsume r s from a ct i ve l y seek i ng subst itutes

2 8 G i ve n the sma l l proport i on of co llege gr a du ates that each y e a rs med i c a l schoo l c l a s s rep r e s e nts th i s seems a re a s on a b l e a s s umpt i on

- 2 8 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

to ch an g i ng l i ce ns i ng laws and imp l i e s an i nc r e a s e of 4000 FMGs

at any g i ve n i ncome le ve l betw e e n 1 9 6 8 and 19 7 5 F i na l l y when

the imm i grat i o n law was ti ghtened i n 1 9 7 7 th e supp l y cu rve

s h i fted ba c k but not a l l th e way to its or i g i na l pos it i on The

1 9 8 1 FMG supp ly cu rve l i e s about hal fway betwe e n its late 1 9 60s

a n d ea r ly - 1 9 70s pos it i on s 2 5

Re s u lt s f rom est imat i n g the unre str i cte d redu c e d form

equat ion ( 6 ) a re pre s ented in co lumn 3 of Tab l e 1 Med i c a l

s choo l constru ct i o n fund i ng ( CONS 8 ) i s used to rep r e s e nt the

dome st i c f r i nge su bs i dy CAPS 5 was al s o used i n some re gre s s i on s

but its h i gh co rre l at i on w ith CONS 8 ( 9 1 ) a s we l l a s w ith th e

F MG re str i ct i ons created prob lems in i d e nt i f y i ng coe f f i c i e nt

e st imates The re g re s s ion was est imated us i ng a Pra i s -W i nsten

tra ns f ormat i on to cor r e ct autocorre l at i on Wh i le it is

i mpos s i b l e to det e rm i ne the net e f f e ct s of th e re str i ct i on

var i a b l e s f rom the theoret i c a l mod e l be c a u s e of the of f s ett i ng

i mp acts of the i r i n f luences on phy s i c i a n i ncome the emp i r i c a l

r e s u lts show at le a st the i r s i g ns ( a nd often the i r coe f f i cient s )

to be s im i l a r to the 2SLS re su lts

F i na l ly we can comp a re what ou r va r i ou s est i mates ind i cate

h as bee n the cumu l at i ve e f f e ct of ch anges in the l i ce n s i ng and

25 The t ime path der i ve d us i ng the unreport e d D6 8 reg re s s i on is s im i l a r exc ept that the outwa rd exp ans i o n be g i n s ea r l i e r and i s mor e gradua l

- 2 4 shy

imm i g r at i on law s s i nc e 1 9 6 5 We me a s u re the ef f e ct of each va r i shy

a b l e i n yea r t by tak i ng the d i f f e rence betwe e n its actu a l ma gn i shy

tude in yea r t and th e va l u e it had in 1 9 6 5 t i me s its est i mated

coe f f i c i e nt Table 2 shows th e cumu lat i v e ef f e ct from 19 6 6

through 1 9 8 1 of th e l i c ens i n g and imm i grat i on reg u l at i ons

e f f e ct s for each of ou r re gres s i ons The re su lt s su ggest that

re l a xat i o n of the s e law s ena b l e d somewh e re f r om 3 3 000 to 4 5 000

add it i ona l FMG s to obta i n us me d i c a l l i c e n s e s du r i ng th e

f i fteen yea r pe r i od The equat i on s suggest a les s e r e f f e ct f rom

the l i c e n s i ng va r i a b l e s equa l i ng 12 000 - 2 2 000 and a net pos ishy

t i ve impact ( rememb e r that the 1 9 7 6 law re str i cted ent ry aga i n )

f rom imm i g r at i on po l i cy of 2 1 000 - 2 4 000 The D7 l imm i g r at i on

va r i ab le a l l owed about 3 000 entra nt s annu a l ly wh i l e D7 6 w ithhe l d

about 2 000 O f the l i ce n s i ng va r i ab le s CITIZ is most i n f l u e nshy

t i a l it ac cou nt s for 60 to 7 5 per c e nt of the tota l l i ce n s i ng

e f f ect The tota l ef f ect s i nc e 1 9 6 5 of loo s e n e d reg u l at i o n s on

FMG ent ry ha s bee n substant i a l equa l i ng 4 5 to 6 5 pe r c e nt of the

app rox i mate l y 70 000 FMGs th at have be come l i c e n s e d i n the us

s i nce 1 9 6 5

P roba b l y the pr i ma ry imp etu s for re l a xat i on of the

Imm i g r a nt and V i s a law s came f rom the pe rce i ve d shortage of

p hy s i c i a ns in the us in the 1 9 60s Moreove r the large

e xp a ns i on of hos p ita l i nte rn s h ip and re s i de ncy tra i n i ng pos it ions

f o l low i ng Wor l d Wa r II pa rt ly du e to the H i l l -Bu rton hosp ita l

constru ct i on prog ram created a dema n d for gra du ate me d i ca l

student s wh i ch th e us schoo l s grad u ates d i d not f i l l The AMA

- 2 5 shy

TAB LE 2

IMPACT OF RELAXATION OF FMG ENTRY RE STRICTIONS FROM 1 9 6 5 LEVEL ON NUMB E R OF NEW FMG LICENSE E S

1 9 6 6 - 1 9 8 1 CUMU LATIVE EFF ECT

2 SLS Redu ced

Form

L i ce ns i ng

Imm i g r at i on

Tota l

1 3 2 7 3

2 1 3 10

3 4 5 8 3

2 1 9 40

2 3 5 1 6

4 5 4 5 6

- 2 6 shy

has certa i n ly advo cated exc lus i on a ry po l i c i es It opposed

r e l a xat i on of the imm i grat i on law in the 1 9 60 s and ba ck in

1 9 3 8 th e AMA House of De l e g ates res o l ve d th at us cit ize ns h ip

be i n c lu ded as a prerequ is ite i n all state me d i ca l l i cens i ng

statutes 2 6 The t i ghten i ng of the imm i grat i on law i n 1 9 7 6 came

at a time when U S schoo ls we re ge nerat i ng a substant i a l ly

l a rge r number of annu a l gra du ates th an they ha d be e n at the

beg i nn i ng of the de cade More over re i mpos it i on of str i ct state

l i c e ns i ng laws has not oc curre d Such act i on is perhaps too

cost ly for th e AMA requ i r i ng 50 sep a rate lobby i ng ef forts

The de c l i ne i n FMG ent ry in the last few yea rs does not

i mp ly that compet it i o n i n the med i ca l prof ess i on has abated

u s me d i ca l scho o ls dou b l e d the i r annu a l output betwe e n 1 9 6 5 and

1 9 80 and in the last few yea rs have mo re tha n ma de up for th e

reduced numb e r of FMG s obta i n i ng us l i c e nses In the ne xt

sect i on we w i l l exam i ne th e det e rm i nants of us new l i ce nsees

i n ord e r to iso lat e th e domest i c compet it i ve f r i nge

III DOMESTIC COMP ETITIVE FRINGE ENTRY

M easur i ng the second comp et it i ve f r i nge th e add it i ona l us

g raduates resu lt i ng so le ly f rom fede ra l ly su bs i d ized me d i c a l

school growth is more comp l i c ated To att r i bute al l o f the 100

p e r c e nt growth betwe e n 1 9 6 5 and 19 80 in the annu a l numb e r of

2 6 I re n e Butte r and Rebe c c a G Sweet L i c e nsu re of Fore i gn M ed i c a l Graduates An H istor i ca l Perspe ct i ve 5 5 Mi l b a nk M emor i a l Fund Qua rt e r ly 3 1 5 ( 1 9 7 7 )

- 2 7 shy

graduates to an ent ry ba rr i e r bre a k down is su re ly an ove rstate shy

ment The adve nt of the Me d i c a i d and Med i c a r e progr ams in 1 9 6 5

i nc r e a s e d dema nd as d i d h i g h e r consume r i ncome s unt i l th e late

1 9 70s There f ore us i ng the dom i nant f i rm mode l pa rt of the

i nc r e a s e in me d i ca l s choo l output ca n proba b l y be att r i buted to

the opt i ma l ad j u stme nt of a ca rte l to growth in dema nd2 7 S i nce

d emand cond it i ons changed after 19 6 5 we ca nnot s imp ly pro j e ct

g rowth in the dom i na nt f i rm from its pre - 1 9 6 5 be h a v i or

If we pos it the ma rg i na l cost curve for phy s i c i a ns as hor i shy

zont a l at the le ve l of tot a l opp ortu n ity co st up to a constra i nt

2 8d et e rm i ne d by th e numbe r of me d i c a l schoo l sp a c e s and ve rt i ca l

the r e a ft e r a growth i n th e dome st i c compet it i ve f r i nge imp l i e s a

r i ghtw ard sh i f t of th e schoo l cap a c i ty constra i nt That is in

t h i s cas e ne ither a lowe r i ng of the opportu n ity cost nor a ri s e

i n phy s i c i a n i ncome ca n le ad to incr e a s e d US gr aduate ent ry

u n l e s s the med i ca l s choo l cap a c ity is increas e d Conve rs e l y th e

a ctu a l re l a xat i on of th e entry ba r r i e r doe s not le a d to a reduced

opp ortu n ity cost for thos e ac c epted i nto me d i c a l schoo l ie

the hor izont a l segme nt of the cu rve is not lowe red

2 7 It is not c l e a r that an ex i st i ng numbe r of ( p e r f ect l y c a rte l ized ) phy s i c i a ns wou l d want to permit entry o f new p hy s i c i a ns rather th an exp a n d i ng the i r ow n output Such a de s i re c a n be rat i on a l ized as an attempt to pre ve nt co nsume r s from a ct i ve l y seek i ng subst itutes

2 8 G i ve n the sma l l proport i on of co llege gr a du ates that each y e a rs med i c a l schoo l c l a s s rep r e s e nts th i s seems a re a s on a b l e a s s umpt i on

- 2 8 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

imm i g r at i on law s s i nc e 1 9 6 5 We me a s u re the ef f e ct of each va r i shy

a b l e i n yea r t by tak i ng the d i f f e rence betwe e n its actu a l ma gn i shy

tude in yea r t and th e va l u e it had in 1 9 6 5 t i me s its est i mated

coe f f i c i e nt Table 2 shows th e cumu lat i v e ef f e ct from 19 6 6

through 1 9 8 1 of th e l i c ens i n g and imm i grat i on reg u l at i ons

e f f e ct s for each of ou r re gres s i ons The re su lt s su ggest that

re l a xat i o n of the s e law s ena b l e d somewh e re f r om 3 3 000 to 4 5 000

add it i ona l FMG s to obta i n us me d i c a l l i c e n s e s du r i ng th e

f i fteen yea r pe r i od The equat i on s suggest a les s e r e f f e ct f rom

the l i c e n s i ng va r i a b l e s equa l i ng 12 000 - 2 2 000 and a net pos ishy

t i ve impact ( rememb e r that the 1 9 7 6 law re str i cted ent ry aga i n )

f rom imm i g r at i on po l i cy of 2 1 000 - 2 4 000 The D7 l imm i g r at i on

va r i ab le a l l owed about 3 000 entra nt s annu a l ly wh i l e D7 6 w ithhe l d

about 2 000 O f the l i ce n s i ng va r i ab le s CITIZ is most i n f l u e nshy

t i a l it ac cou nt s for 60 to 7 5 per c e nt of the tota l l i ce n s i ng

e f f ect The tota l ef f ect s i nc e 1 9 6 5 of loo s e n e d reg u l at i o n s on

FMG ent ry ha s bee n substant i a l equa l i ng 4 5 to 6 5 pe r c e nt of the

app rox i mate l y 70 000 FMGs th at have be come l i c e n s e d i n the us

s i nce 1 9 6 5

P roba b l y the pr i ma ry imp etu s for re l a xat i on of the

Imm i g r a nt and V i s a law s came f rom the pe rce i ve d shortage of

p hy s i c i a ns in the us in the 1 9 60s Moreove r the large

e xp a ns i on of hos p ita l i nte rn s h ip and re s i de ncy tra i n i ng pos it ions

f o l low i ng Wor l d Wa r II pa rt ly du e to the H i l l -Bu rton hosp ita l

constru ct i on prog ram created a dema n d for gra du ate me d i ca l

student s wh i ch th e us schoo l s grad u ates d i d not f i l l The AMA

- 2 5 shy

TAB LE 2

IMPACT OF RELAXATION OF FMG ENTRY RE STRICTIONS FROM 1 9 6 5 LEVEL ON NUMB E R OF NEW FMG LICENSE E S

1 9 6 6 - 1 9 8 1 CUMU LATIVE EFF ECT

2 SLS Redu ced

Form

L i ce ns i ng

Imm i g r at i on

Tota l

1 3 2 7 3

2 1 3 10

3 4 5 8 3

2 1 9 40

2 3 5 1 6

4 5 4 5 6

- 2 6 shy

has certa i n ly advo cated exc lus i on a ry po l i c i es It opposed

r e l a xat i on of the imm i grat i on law in the 1 9 60 s and ba ck in

1 9 3 8 th e AMA House of De l e g ates res o l ve d th at us cit ize ns h ip

be i n c lu ded as a prerequ is ite i n all state me d i ca l l i cens i ng

statutes 2 6 The t i ghten i ng of the imm i grat i on law i n 1 9 7 6 came

at a time when U S schoo ls we re ge nerat i ng a substant i a l ly

l a rge r number of annu a l gra du ates th an they ha d be e n at the

beg i nn i ng of the de cade More over re i mpos it i on of str i ct state

l i c e ns i ng laws has not oc curre d Such act i on is perhaps too

cost ly for th e AMA requ i r i ng 50 sep a rate lobby i ng ef forts

The de c l i ne i n FMG ent ry in the last few yea rs does not

i mp ly that compet it i o n i n the med i ca l prof ess i on has abated

u s me d i ca l scho o ls dou b l e d the i r annu a l output betwe e n 1 9 6 5 and

1 9 80 and in the last few yea rs have mo re tha n ma de up for th e

reduced numb e r of FMG s obta i n i ng us l i c e nses In the ne xt

sect i on we w i l l exam i ne th e det e rm i nants of us new l i ce nsees

i n ord e r to iso lat e th e domest i c compet it i ve f r i nge

III DOMESTIC COMP ETITIVE FRINGE ENTRY

M easur i ng the second comp et it i ve f r i nge th e add it i ona l us

g raduates resu lt i ng so le ly f rom fede ra l ly su bs i d ized me d i c a l

school growth is more comp l i c ated To att r i bute al l o f the 100

p e r c e nt growth betwe e n 1 9 6 5 and 19 80 in the annu a l numb e r of

2 6 I re n e Butte r and Rebe c c a G Sweet L i c e nsu re of Fore i gn M ed i c a l Graduates An H istor i ca l Perspe ct i ve 5 5 Mi l b a nk M emor i a l Fund Qua rt e r ly 3 1 5 ( 1 9 7 7 )

- 2 7 shy

graduates to an ent ry ba rr i e r bre a k down is su re ly an ove rstate shy

ment The adve nt of the Me d i c a i d and Med i c a r e progr ams in 1 9 6 5

i nc r e a s e d dema nd as d i d h i g h e r consume r i ncome s unt i l th e late

1 9 70s There f ore us i ng the dom i nant f i rm mode l pa rt of the

i nc r e a s e in me d i ca l s choo l output ca n proba b l y be att r i buted to

the opt i ma l ad j u stme nt of a ca rte l to growth in dema nd2 7 S i nce

d emand cond it i ons changed after 19 6 5 we ca nnot s imp ly pro j e ct

g rowth in the dom i na nt f i rm from its pre - 1 9 6 5 be h a v i or

If we pos it the ma rg i na l cost curve for phy s i c i a ns as hor i shy

zont a l at the le ve l of tot a l opp ortu n ity co st up to a constra i nt

2 8d et e rm i ne d by th e numbe r of me d i c a l schoo l sp a c e s and ve rt i ca l

the r e a ft e r a growth i n th e dome st i c compet it i ve f r i nge imp l i e s a

r i ghtw ard sh i f t of th e schoo l cap a c i ty constra i nt That is in

t h i s cas e ne ither a lowe r i ng of the opportu n ity cost nor a ri s e

i n phy s i c i a n i ncome ca n le ad to incr e a s e d US gr aduate ent ry

u n l e s s the med i ca l s choo l cap a c ity is increas e d Conve rs e l y th e

a ctu a l re l a xat i on of th e entry ba r r i e r doe s not le a d to a reduced

opp ortu n ity cost for thos e ac c epted i nto me d i c a l schoo l ie

the hor izont a l segme nt of the cu rve is not lowe red

2 7 It is not c l e a r that an ex i st i ng numbe r of ( p e r f ect l y c a rte l ized ) phy s i c i a ns wou l d want to permit entry o f new p hy s i c i a ns rather th an exp a n d i ng the i r ow n output Such a de s i re c a n be rat i on a l ized as an attempt to pre ve nt co nsume r s from a ct i ve l y seek i ng subst itutes

2 8 G i ve n the sma l l proport i on of co llege gr a du ates that each y e a rs med i c a l schoo l c l a s s rep r e s e nts th i s seems a re a s on a b l e a s s umpt i on

- 2 8 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

TAB LE 2

IMPACT OF RELAXATION OF FMG ENTRY RE STRICTIONS FROM 1 9 6 5 LEVEL ON NUMB E R OF NEW FMG LICENSE E S

1 9 6 6 - 1 9 8 1 CUMU LATIVE EFF ECT

2 SLS Redu ced

Form

L i ce ns i ng

Imm i g r at i on

Tota l

1 3 2 7 3

2 1 3 10

3 4 5 8 3

2 1 9 40

2 3 5 1 6

4 5 4 5 6

- 2 6 shy

has certa i n ly advo cated exc lus i on a ry po l i c i es It opposed

r e l a xat i on of the imm i grat i on law in the 1 9 60 s and ba ck in

1 9 3 8 th e AMA House of De l e g ates res o l ve d th at us cit ize ns h ip

be i n c lu ded as a prerequ is ite i n all state me d i ca l l i cens i ng

statutes 2 6 The t i ghten i ng of the imm i grat i on law i n 1 9 7 6 came

at a time when U S schoo ls we re ge nerat i ng a substant i a l ly

l a rge r number of annu a l gra du ates th an they ha d be e n at the

beg i nn i ng of the de cade More over re i mpos it i on of str i ct state

l i c e ns i ng laws has not oc curre d Such act i on is perhaps too

cost ly for th e AMA requ i r i ng 50 sep a rate lobby i ng ef forts

The de c l i ne i n FMG ent ry in the last few yea rs does not

i mp ly that compet it i o n i n the med i ca l prof ess i on has abated

u s me d i ca l scho o ls dou b l e d the i r annu a l output betwe e n 1 9 6 5 and

1 9 80 and in the last few yea rs have mo re tha n ma de up for th e

reduced numb e r of FMG s obta i n i ng us l i c e nses In the ne xt

sect i on we w i l l exam i ne th e det e rm i nants of us new l i ce nsees

i n ord e r to iso lat e th e domest i c compet it i ve f r i nge

III DOMESTIC COMP ETITIVE FRINGE ENTRY

M easur i ng the second comp et it i ve f r i nge th e add it i ona l us

g raduates resu lt i ng so le ly f rom fede ra l ly su bs i d ized me d i c a l

school growth is more comp l i c ated To att r i bute al l o f the 100

p e r c e nt growth betwe e n 1 9 6 5 and 19 80 in the annu a l numb e r of

2 6 I re n e Butte r and Rebe c c a G Sweet L i c e nsu re of Fore i gn M ed i c a l Graduates An H istor i ca l Perspe ct i ve 5 5 Mi l b a nk M emor i a l Fund Qua rt e r ly 3 1 5 ( 1 9 7 7 )

- 2 7 shy

graduates to an ent ry ba rr i e r bre a k down is su re ly an ove rstate shy

ment The adve nt of the Me d i c a i d and Med i c a r e progr ams in 1 9 6 5

i nc r e a s e d dema nd as d i d h i g h e r consume r i ncome s unt i l th e late

1 9 70s There f ore us i ng the dom i nant f i rm mode l pa rt of the

i nc r e a s e in me d i ca l s choo l output ca n proba b l y be att r i buted to

the opt i ma l ad j u stme nt of a ca rte l to growth in dema nd2 7 S i nce

d emand cond it i ons changed after 19 6 5 we ca nnot s imp ly pro j e ct

g rowth in the dom i na nt f i rm from its pre - 1 9 6 5 be h a v i or

If we pos it the ma rg i na l cost curve for phy s i c i a ns as hor i shy

zont a l at the le ve l of tot a l opp ortu n ity co st up to a constra i nt

2 8d et e rm i ne d by th e numbe r of me d i c a l schoo l sp a c e s and ve rt i ca l

the r e a ft e r a growth i n th e dome st i c compet it i ve f r i nge imp l i e s a

r i ghtw ard sh i f t of th e schoo l cap a c i ty constra i nt That is in

t h i s cas e ne ither a lowe r i ng of the opportu n ity cost nor a ri s e

i n phy s i c i a n i ncome ca n le ad to incr e a s e d US gr aduate ent ry

u n l e s s the med i ca l s choo l cap a c ity is increas e d Conve rs e l y th e

a ctu a l re l a xat i on of th e entry ba r r i e r doe s not le a d to a reduced

opp ortu n ity cost for thos e ac c epted i nto me d i c a l schoo l ie

the hor izont a l segme nt of the cu rve is not lowe red

2 7 It is not c l e a r that an ex i st i ng numbe r of ( p e r f ect l y c a rte l ized ) phy s i c i a ns wou l d want to permit entry o f new p hy s i c i a ns rather th an exp a n d i ng the i r ow n output Such a de s i re c a n be rat i on a l ized as an attempt to pre ve nt co nsume r s from a ct i ve l y seek i ng subst itutes

2 8 G i ve n the sma l l proport i on of co llege gr a du ates that each y e a rs med i c a l schoo l c l a s s rep r e s e nts th i s seems a re a s on a b l e a s s umpt i on

- 2 8 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

has certa i n ly advo cated exc lus i on a ry po l i c i es It opposed

r e l a xat i on of the imm i grat i on law in the 1 9 60 s and ba ck in

1 9 3 8 th e AMA House of De l e g ates res o l ve d th at us cit ize ns h ip

be i n c lu ded as a prerequ is ite i n all state me d i ca l l i cens i ng

statutes 2 6 The t i ghten i ng of the imm i grat i on law i n 1 9 7 6 came

at a time when U S schoo ls we re ge nerat i ng a substant i a l ly

l a rge r number of annu a l gra du ates th an they ha d be e n at the

beg i nn i ng of the de cade More over re i mpos it i on of str i ct state

l i c e ns i ng laws has not oc curre d Such act i on is perhaps too

cost ly for th e AMA requ i r i ng 50 sep a rate lobby i ng ef forts

The de c l i ne i n FMG ent ry in the last few yea rs does not

i mp ly that compet it i o n i n the med i ca l prof ess i on has abated

u s me d i ca l scho o ls dou b l e d the i r annu a l output betwe e n 1 9 6 5 and

1 9 80 and in the last few yea rs have mo re tha n ma de up for th e

reduced numb e r of FMG s obta i n i ng us l i c e nses In the ne xt

sect i on we w i l l exam i ne th e det e rm i nants of us new l i ce nsees

i n ord e r to iso lat e th e domest i c compet it i ve f r i nge

III DOMESTIC COMP ETITIVE FRINGE ENTRY

M easur i ng the second comp et it i ve f r i nge th e add it i ona l us

g raduates resu lt i ng so le ly f rom fede ra l ly su bs i d ized me d i c a l

school growth is more comp l i c ated To att r i bute al l o f the 100

p e r c e nt growth betwe e n 1 9 6 5 and 19 80 in the annu a l numb e r of

2 6 I re n e Butte r and Rebe c c a G Sweet L i c e nsu re of Fore i gn M ed i c a l Graduates An H istor i ca l Perspe ct i ve 5 5 Mi l b a nk M emor i a l Fund Qua rt e r ly 3 1 5 ( 1 9 7 7 )

- 2 7 shy

graduates to an ent ry ba rr i e r bre a k down is su re ly an ove rstate shy

ment The adve nt of the Me d i c a i d and Med i c a r e progr ams in 1 9 6 5

i nc r e a s e d dema nd as d i d h i g h e r consume r i ncome s unt i l th e late

1 9 70s There f ore us i ng the dom i nant f i rm mode l pa rt of the

i nc r e a s e in me d i ca l s choo l output ca n proba b l y be att r i buted to

the opt i ma l ad j u stme nt of a ca rte l to growth in dema nd2 7 S i nce

d emand cond it i ons changed after 19 6 5 we ca nnot s imp ly pro j e ct

g rowth in the dom i na nt f i rm from its pre - 1 9 6 5 be h a v i or

If we pos it the ma rg i na l cost curve for phy s i c i a ns as hor i shy

zont a l at the le ve l of tot a l opp ortu n ity co st up to a constra i nt

2 8d et e rm i ne d by th e numbe r of me d i c a l schoo l sp a c e s and ve rt i ca l

the r e a ft e r a growth i n th e dome st i c compet it i ve f r i nge imp l i e s a

r i ghtw ard sh i f t of th e schoo l cap a c i ty constra i nt That is in

t h i s cas e ne ither a lowe r i ng of the opportu n ity cost nor a ri s e

i n phy s i c i a n i ncome ca n le ad to incr e a s e d US gr aduate ent ry

u n l e s s the med i ca l s choo l cap a c ity is increas e d Conve rs e l y th e

a ctu a l re l a xat i on of th e entry ba r r i e r doe s not le a d to a reduced

opp ortu n ity cost for thos e ac c epted i nto me d i c a l schoo l ie

the hor izont a l segme nt of the cu rve is not lowe red

2 7 It is not c l e a r that an ex i st i ng numbe r of ( p e r f ect l y c a rte l ized ) phy s i c i a ns wou l d want to permit entry o f new p hy s i c i a ns rather th an exp a n d i ng the i r ow n output Such a de s i re c a n be rat i on a l ized as an attempt to pre ve nt co nsume r s from a ct i ve l y seek i ng subst itutes

2 8 G i ve n the sma l l proport i on of co llege gr a du ates that each y e a rs med i c a l schoo l c l a s s rep r e s e nts th i s seems a re a s on a b l e a s s umpt i on

- 2 8 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

graduates to an ent ry ba rr i e r bre a k down is su re ly an ove rstate shy

ment The adve nt of the Me d i c a i d and Med i c a r e progr ams in 1 9 6 5

i nc r e a s e d dema nd as d i d h i g h e r consume r i ncome s unt i l th e late

1 9 70s There f ore us i ng the dom i nant f i rm mode l pa rt of the

i nc r e a s e in me d i ca l s choo l output ca n proba b l y be att r i buted to

the opt i ma l ad j u stme nt of a ca rte l to growth in dema nd2 7 S i nce

d emand cond it i ons changed after 19 6 5 we ca nnot s imp ly pro j e ct

g rowth in the dom i na nt f i rm from its pre - 1 9 6 5 be h a v i or

If we pos it the ma rg i na l cost curve for phy s i c i a ns as hor i shy

zont a l at the le ve l of tot a l opp ortu n ity co st up to a constra i nt

2 8d et e rm i ne d by th e numbe r of me d i c a l schoo l sp a c e s and ve rt i ca l

the r e a ft e r a growth i n th e dome st i c compet it i ve f r i nge imp l i e s a

r i ghtw ard sh i f t of th e schoo l cap a c i ty constra i nt That is in

t h i s cas e ne ither a lowe r i ng of the opportu n ity cost nor a ri s e

i n phy s i c i a n i ncome ca n le ad to incr e a s e d US gr aduate ent ry

u n l e s s the med i ca l s choo l cap a c ity is increas e d Conve rs e l y th e

a ctu a l re l a xat i on of th e entry ba r r i e r doe s not le a d to a reduced

opp ortu n ity cost for thos e ac c epted i nto me d i c a l schoo l ie

the hor izont a l segme nt of the cu rve is not lowe red

2 7 It is not c l e a r that an ex i st i ng numbe r of ( p e r f ect l y c a rte l ized ) phy s i c i a ns wou l d want to permit entry o f new p hy s i c i a ns rather th an exp a n d i ng the i r ow n output Such a de s i re c a n be rat i on a l ized as an attempt to pre ve nt co nsume r s from a ct i ve l y seek i ng subst itutes

2 8 G i ve n the sma l l proport i on of co llege gr a du ates that each y e a rs med i c a l schoo l c l a s s rep r e s e nts th i s seems a re a s on a b l e a s s umpt i on

- 2 8 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

Us i ng th e mode l out l i ne d i n sect i on I we ca n de r i ve an

equat i on that measu res e f f e cts of go ve rnme nt subs i d i es to me d i c a l

s choo ls on the numb e r of annu a l new domest i c l i ce nses We are

v i ew i ng the fede r a l gra nts as ne gat i ve restr i ct i ons a sort of

e ntry induceme nt rather th an entry ba rr i e r The poo l of total

new domest i c l i ce nsees ( DNL ) is composed of two grou ps ( 1 ) a

domest i c compet it i ve f r i nge ( CDNL ) and ( 2 ) the opt ima l ad j ustment

of the ca rte l to cha nges in dema nd or cost ( DFNL ) As w ith the

f ore i gn f r i n ge sup p l y cu rve entry of the domest i c f r i nge ca n be

desc r i be d as

=( 7 ) CDNLt d + d MDINC + dRDt + d TOCt0 1 t - 2-- 3

where d gt 0 d lt 0 and d lt 0 si nce RD equ a ls the ve ctor of 1 2 3

domest i c restr i ct i ons and for th e domest i c f r i nge FTOC is just

the us phys i c i a n s opportu n ity cost along the hor izont a l

segme nt

The RD ve ctor is me asu r e d as the subs i d i es g i ve n to me d i c a l

s choo ls these ne gat i ve restr i ct i ons exp a nd schoo l cap a c ity and

w ea k e n the hypothes ized domest i c ent ry ba r r i e r enab l i ng de ve lop shy

ment of the f r i nge 2 9 F e de r a l gra nts to med i c a l schoo ls were

In ou r mode l of TOC they sh i ft to the r i ght the ve rt i ca l segme nt of the cost cu r ve That is they wea k e n the AMA s a b i l ity to constra i n me d i c a l schoo l spa ces I n th at they may i nd i rect ly af f e ct tu it i on fees th ey also se rve to lower the h or izont a l segme nt T h is imp a ct is measu red i n TOC and the r e shyf ore ou r est i mated coe f f i c i e nts o n the schoo l subs i dy va r i ables w i l l be bi as e d downwa rds We also do not inc lude the ef f e cts of qu ite su bsta nt i a l state su bs i d i es to med i c a l schoo ls

- 2 9 shy

29

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

d i str i buted unde r two di f f e re nt progr ams B eg i n n i ng in 1 9 6 5

construct i on funds ( CON S ) we re g i ve n for exp a n s i o n and bu i l d i ng

o f new fac i l it i es In 1 9 6 6 cap itat i on fund i ng ( CAPS ) wa s also

i n it i ated to be e l i g i b le for th e s e gra nts schoo l s we re requ i re d

t o expand en rol lme nt annu a l ly by 2 12 percent o r f i ve stu dents

w h i chever w a s gre at e r 30 S i nce con struct i on fu n d s have a

p e rma ne nt impa ct ( e xcept for dep re c i at i on ) on med i c a l schoo l

c ap a c ity th ey are entered as the cumu l at i ve sum i n emp i r i c a l

t ests Moreove r both va r i ab le s shou ld af f e ct new l i ce n s e e s wit h

a lag cap itat i on funds are pres ume d to tak e f i ve yea rs g i ve n

t h e nec e s s ary tra i n i ng t ime wh i le co nstru ct i on fu nds are lagged

f or e i ght yea r s as s u m i n g the same f i ve year tra i n i ng hor izon

p lu s an add it i on a l th ree for actu a l constru ct i on 3 1

Equat ion ( 1 ) rep r e s e nt s the domi na nt f i rm port i on of new

d omest i c li c e n s e e s 3 2 who are assume d to ad j u st pa rt i a l ly to

e xpected equ i l i br i um stoc k cha n g e s net of f r i nge ( both fore i gn

a nd dome st i c ) entry and adapt fu l ly to the k now n and constant

depre c i at i on rate d Rew r it i ng ( 1 ) to sep a r ate the domest i c and

f ore i gn f r i nges y i e l d s

30 us Hou se of Rep re s e nt at i ve s Comm i tt e e on Int e r state and F ore i gn Comme rce Cu r r e nt Hea lt h Manpower Is s u e s at 6- 8

3 1 Seven and n i ne year la g s were tr i e d as we l l and produced s im i l a r re s u lt s

3 2 S i nce a l l FMGs are as sume d to be part of the f r i nge al l of the dom i nant f i rm is dome s t i c

- 30 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

( MDSt - l l t - 5 + FMGt l t - 5 + CDNLt l t - 5 ) ]( 8 ) DFNLt = y[MDSt l t - 5

+ dMDSt - l l t - 5bull

Summ i ng equat i ons ( 7 ) and ( 8 ) and mak i ng app r op r i ate subst itu shy

t i ons le ads to the f o l low i ng est i mat a b le equat i on me as u r i ng

tot a l domest i c ent ry

4 + n 8FMGPt l t - 5 + Et

where DNLP is the dome st i c new l i ce n s e e -popu lat i on rat i o

Th i s is est imat e d as a 2SLS re gr e s s i on w ith ( 5 ) to mea su r e

MDINc 3 3 n 1 n 3 n 4 n 5 a n d n 8 s hou ld be ne gat i ve wh i l e n 2 1r

n 2 2r n 6 lr n 6 2 and n 7 s hou ld be pos it i ve

We a l s o de ve lop the re duced form to de r i ve the net ef f ect of

s choo l subs i d i e s Th i s inc ludes not on ly the di rect ef f e ct on

the compet it i ve f r i nge but its of f s ett i n g impa ct du e to its

negat i ve pu l l on i ncome and the re su ltant d i r e ct inf luence due to

lower i ncome s ne gat i ve e f f e ct on FMG entry We est i mate the

reduced form as

3 3 FMGt l t - 5 is est i mate d us i n g a Box -J enk i ns f i rst-order autoshyregre s s i ve f i ve ye a r for e c a st CITI Z is used to rep re s e nt the re str i ct i on det e rm i na nt s of FMG entry in the dome st i c l i ce n s e e regre s s i on

- 3 1 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

-FMGt i t - 5 middot Box-J enk i ns

=( 10 ) DNLPt mo + m lTOCt + m2 1 INCMt + m2 2 PUBINSt + m3CITI Z t

F f + m4MDSPt - l + m5et + m6 1CON S 8t + m6 2CAP S 5t + m7et

5 3 4 + Et bull

Res u lts from the 2SLS structu r a l mode l regre s s i on are

p r e s e nted in Co lumn 1 of Tab l e 3 The income equat i on that forms

the pre d i ct i ons is ide nt i c a l to that used for FMG s and reporte d

i n Ta b le 2 It is not rep e at e d he re

TOC INCM and CITI Z have pre d i cted ef f e ct s and are ma i n ly

s i gn i f i c a nt at the 0 2 5 le ve l PUBINS shou l d be pos it i ve but is

Fnegat i ve inste a d The coe f f i c i e nt s on MDSPt - 1 and e a re con shy

stra i ne d to sum to zero they ea ch tak e the app r op r i ate s i gns but

a re ins i gn i f i c a nt P re d i cted MDINC is cor r e ct ly pos it i ve but not

at al l s i gn i f i c a nt The f i ve-y e a r FMG for e c a s t is als o tota l ly

i ns i g n i f i c a nt In l i eu of the Box-Jenk i ns est i mated FMGt l t - 5 we

a l s o used th e FMG pre d i ct i on ge nerated by the est i mat i on of ( 4 )

i n th e 2SLS re gr e s s i on di s cu s s e d i n the la st sect i on The

r e s u lt s we re ve ry s im i l a r and he nce are not report e d

CON S 8 and CAPS 5 have a pre d i cted pos it i ve ef f ect Ne ith er

a re ove r ly s i gn i f i c a nt howeve r th ey are h i g h ly co rre lated w ith

f repre s e nt s the pre d i ct i on error on yea r ts actu a l FMGet

e nt ry i e FMGt I n it i a l use of a proxy for F MGPt l t - 5 ba s e d on a mode l le d to an i napprop r i shy

f ate ly ne gat i ve and s i gn i f i ca nt coe f f i c i e nt on etbull Th i s su ggests that the exp e ct a -t i on s proce s s was incorrect ly mode l e d No e f term is used i n the est imat i o n s report e d

- 3 2 shy

3 4

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

------------

TAB LE 3

ESTIMAT ES OF NEW DOME STIC LIC E N S E E EQUATION STRUCTURAL ( 2SLS ) ESTIMATES ( 9 ) AND RE DUCE D FORM ESTIMAT ES ( 10 )

1 9 4 6 - 1 9 8 1

( 1 ) Stru ctu ra l

INTERCEPT 2 9 1 3 ( - 1 ) ( 1 0 3 )

TOC - 3 40 1 ( - 2 ) ( - 2 5 6 )

INCM 2 2 3 5 ( - 1 ) ( 2 5 9 )

PUB INS - 3 9 5 4 ( - 3 ) ( - 1 2 6 )

MDSP ( t - 1 ) - 7 7 9 7 ( - 2 ) ( - 5 2 )

F 7 7 9 7 ( - 2 ) ( bull 5 2 )

CITIZ 2 3 8 3 ( - 3 )

e

( 2 0 6 )

CON S 8 1 7 8 9 ( - 1 ) ( 9 2 )

CAPS 5 50 8 ( - 1 ) ( 1 1 5 )

MD INC 8 3 3 4 ( - 3 ) ( 7 2 )

FMGPt t - 5 1 3 3 5 ( - 1 ) ( bull 0 9 )

S S E 1 2 7 6 ( - 3 )

R 2 9 6 4 6

D W l 4 2

p

Not e s S ee Tab le 2

Redu c e d Form ( 2 )

9 4 1 1 ( - 2 ) ( 2 8 ) bull

- 2 3 9 6 ( - 2 ) ( - 3 9 2 )

2 4 10 ( - 1 ) ( 3 0 6 )

- 2 9 20 ( - 3 ) ( - 1 3 1 )

2 20 9 ( - 2 ) ( bull 0 9 )

2 4 1 3 ( - 1 ) ( 1 0 1 )

2 6 50 ( - 3 ) ( 2 2 9 )

6 4 90 ( - 2 ) ( 4 8 )

6 7 5 7 ( - 1 ) ( 1 50 )

1 1 7 7 ( - 3 )

9 4 7 3

- 2 5 3

- 3 3 shy

-

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

each oth e r as we l l as w ith oth e r va r i a b l e s in the re gre s s i o n

i nc l u d i ng CITI Z and FMGt l t - 5middot Regre s s i on s that inc luded on ly one

of the schoo l su bs i dy va r i a b l e s or a we i ghted sum of the two wer e

a l s o ru n The re su lt i ng t -stat i st i c s on the schoo l subs i dy va r i shy

a b le s were about 2 The ef f e ct that CON S 8 and CAPS 5 have on DNLP

i so l ates the dome st i c compet it i ve f r i nge The re gre s s i on

s u gge sts a cumu l at i ve ef f e ct from 1 9 7 1 whe n CAPS 5 beg i ns th rou g h

1 9 8 1 of 20 000 add i t i ona l new l i ce n s e e s

To ar r i ve at a tru e net me a s u re of the ef f ect of the me d i c a l

s choo l subs i d i e s we a l s o est i mate th e re duced form of the new

dome st i c l i ce ns e e s supp ly cu rve spec i f i ed by ( 10 ) The data have

a ga i n bee n tra n s f orme d to cor r e ct re s i du a l corre lat i on The s e

resu lts are out l i ne d i n Co lumn 2 of Tab le 3 CITI Z is aga i n used

to mea s u re for e i gn restr i ct i ons They su ggest a cumu l at i ve

dome st i c f r i nge of about 1 3 000 from 1 9 6 6 th rough 1 9 8 1 As

e xpected th i s net f i gure is lower tha n the 20 000 suggested by

the 2 SLS est i mate

IV SUMMARY

In th i s pap e r we have de ve loped a mode l of the phy s i c i an

stock as a dom i na nt f i rm of AMA-sanct i o n e d phy s i c i ans who du e to

va r i ou s gove rnme nt po l i c i e s are faced w ith a grow i n g compet it i ve

f r i nge Two sou r c e s of compet it i on app e a r to ex i st Dur i n g th e

l at e 1 9 60s and ea r l y 1 9 70 s fo l l ow i ng the re la xat i on of va r i ou s

f edera l imm i g r at i on and state l i c e n s i ng rest r i ct i ons fore i gn

- 3 4 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

3 5

me d i c a l graduates forme d a large proport i o n of new us l i ce n shy

s ees As Table 4 shows we est imate that s i nc e 1 9 6 5 3 5 000shy

4 5 000 have be e n able to ent e r be cau se of le s s seve re imm i g r at i o n

a nd l i ce ns i ng re str i ct i ons A dome st i c f r i nge has a l s o ar i s en

p a rt i cu l a r ly in ve ry re c e nt yea rs du e to subs i d i e s gra nted us

med i c a l schoo l s By 1 9 8 1 th i s sou rce of compet it i o n has added

a bout 1 3 000 to 20 000 phy s i c i ans to the act i ve stoc k Thu s ou r

e st i mate s su ggest that an add it i on a l 5 5 000 to 60 000 phy s i c i a ns

o r 1 2 - 1 3 per c e nt of the 19 8 1 act i ve stock pra ct i ce today be cau s e

o f inc r e a s e d compet it i on s i nce 1 9 6 5

The s e est i mat es are con s i stent with tho s e de r i ve d by d i f f e r shy

e nt methods i n a re lated pap e r 3 5 There the actu a l phy s i c i an

stock wa s mode l e d as a we i g hted ave rage of the supp l i es that

wou ld ex i st unde r the ext reme cond it i ons of pu re monopo l y and

p e r f e ct compet it i o n Changes in the de gree of compet it i on we re

mea s u r e d as changes in the we i ght from its 1 9 6 5 le ve l Thos e

e st imates attr i bute 9- 1 4 per c e nt or 39 000 to 6 2 000 of the

1 9 8 1 phy s i c i a n stock to i n c r e a s e d compet it i o n Income i s l i k e shy

w i s e re duced by $ 7 000 - $ 1 1 000 in 1 9 7 2 aft e r tax do l l a r s

I n both th e 2SLS and re duced form est i mate s the i n c r e a s e

due t o the FMG f r i nge i s large r th an that att r i butable t o the

dome st i c compet it i ve f r i nge compr i s i ng 6 2 - 7 8 pe r c e nt of the

tot a l The est i mate s de r i ve d from cumu l at i n g yea r ly entry in the

Noether sup r a note l

- 3 5 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

TAB LE 4

COMPARISON OF DIF F ERENT ESTIMATES OF TH E AMOU NT OF THE 1 9 8 1 PHYSICIAN STOCK AND INCOME DU E TO INCREAS E D

MARKET COMPETITION SINCE 1 9 6 5

M ode l Stock Income

2SLS Est imates

FMG Ef f ect 3 4 5 8 3 2 1 20

Dome st i c E f f ect 20 0 7 4 - 1 2 3 30

Tota l E f f ect 5 4 6 5 7 - 10 2 10

Redu ced Form-Est imat es

FMG E f f e ct 4 5 4 5 6

Dome st i c E f f ect 1 2 8 10

Tot a l E f f ect 5 8 2 6 6 n a

Not e s FMG ef f e ct is the f r i nge est imat e d from the FMG supp ly equ at i on Dome st i c ef f ect is the f r i ng e me a s u re d in the new dome st i c l i c e n s e e equat i on

- 3 6 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

domi nant f i rm mode l are bi a s ed upwards s l i ght ly s i nce dep re c i a shy

t i on has not bee n su b -stra cted from the f i gure s The 2SLS

regre s s i on prov i de s an est i mate of a $ 10 000 income los s from th e

i ncome equat i on s 3 6 No est i mate is de r i ve d in the re duced form

f r i nge mode l

In wh i ch est i mates are we most conf i de nt Theoret i ca l ly

the re duced form equat i on s shou l d gi ve cor r e ct est i mates of the

net ef f e ct of the re l a xat i on of re str i ct i ons on f r i nge ent ry

The net ef f e ct shou ld be sma l l e r tha n that mea s u re d by the 2SLS

p roce s s be c a u s e of of f sett i n g e f f e cts of the re str i ct i on va r i shy

a b les on i ncome wh i c h are incorporated i nto the MDINC term in the

2 SLS est i mat i on Th i s con c l u s i on ho l d s for the new dome st i c

l i ce n s e e cu rve Howeve r i n est i mat i n g the FMG su pp ly cu rve th e

net est imat e of the FMG f r i nge de r i ve d f rom the re duced form

e quat i o n exc e e d s that est i mate d by 2SLs 3 7 The redu c e d form

equat i on doe s have a lower SS E and has be e n cor r e cted for mi ld

autocorr e l at i o n The re duced form est imat e s suggest a tota l

e ntry of compet it i ve f r i nge of ove r 58 000

We ca n a l s o exam i ne the yea r ly patt e rn of entry de r i ve d from

the 2 S LS and redu c e d form est imator s Tab le 5 prov i des th e s e

3 6 The $ 10 2 10 de c l i ne c a n be de comp o s e d i nto a $ 2 1 9 4 lo s s du e to re l a xat i on of FMG l i c e ns i ng restr i ct i on s a $ 4 3 1 4 net ga i n f rom changes i n imm i g r at i on po l i cy and a $ 1 2 000 lo s s du e to med i ca l s choo l subs i d i es The s e numbers g i ve a di f f e re nt vi ew than the stock f i gures wh i c h att r i bute th e ma j o r ity of the growth of compet it i on to the fore i g n me d i ca l graduate f r i n g e

3 7 Th i s i s du e to the CITI Z va r i a b l e i n the D7 1 re gres s i on

- 3 7 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

175

7 3 437

7 4

7 5

76

79

TABLE 5

COMPARISON OF YEARLY ESTIMATES OF COMPETITIVE FRINCE ENTRY 1966-198 1

Reduced Form 2SLS Year Total FMG DNL Total FMG DNL

1 966 0 0 0 0 0 0

67 28 28 0 32 32 0

68 290 290 0 175

69 506 506 0 270 270 0

1970 1 153 1 153 0 638 638 0

7 1 4 350 4 234 1 16 3 577 3 490 87

7 2 4 612 4 28 5 327 3 729 3 483 246

5 186 4 749 4 299 3 778

5 458 4 944 5 14 4 709 3 982

5 811 5 147 662 5 352 4 19 6 1 156

6 043 5 246 797 5 818 4 289 1 529

7 7 4 9 66 3 020 1 946 4 806 2 127 2 679

78 5 245 3 115 2 130 5 353 2 221 3 132

5 214 2 94 1 2 27 3 5 516 2 013 3 503

1980 4 874 2 910 1 9 64 5 282 1 963 3 319

8 1 4 532 2 888 1 644 5 101 1 926 3 175

Total 58 266 4 5 456 12 810 54 657 34 583 20 074

Notes See Table 4

-38shy

0

521

7 27

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

est i mates In both ca s es ent ry of the FMG f r i nge is strongest

in the ea r l y- 1 9 70s and has re ceded i n re c e nt yea rs On th e

oth e r hand the dome st i c compet it i ve fr i nge d i d not ex i st unt i l

1 9 7 1 3 8 and d i d not be come substant i a l unt i l the late 1 9 70 s

The dome st i c f r i nge has also re ceded s l i ght ly in the last two

years as the schoo l su bs i dy programs have bee n phas e d out Th i s

suggest s that a new equ i l i br i um le ve l of compet it i on i s be i ng

approa c h e d

We have shown i n th i s pap e r th at va r i ou s po l i c i e s pro shy

mu l gated by th e AMA we re ef f e ct i ve i n restr i ct i ng the supp ly of

phy s i c i a n s Thu s th e i r re la xat i on ha s led to inc r e a s ed comp et i shy

t i on amo n g phy s i c i an s U na nswere d howeve r is the quest i on of

what factors cau s e d th e po l i cy ch anges that have a ltered the

me d i c a l ma rk et C e rta i n ly th e AMA is one i n f l u e nc i a l actor that

s hou ld be exami ned i n the f ramework of a pol it i c a l support mode l

a s out l i ne d by Pe ltzman 3 9 D i d its powe r wea k e n abs o l utely or

only re lat i ve to the r i nte rest grou p s The answe r to th i s

quest i on i s beyond the scope of th i s pap e r but shou ld provoke

i ntere st i ng futu re work

3 8 The f i ve -yea r la g of the cap itat i on program wh i ch beg a n in 1 9 6 6 accounts for th i s

3 9 Sam Pe ltzman Toward a More Gene ra l Theory of Regu lat i on 1 9 J ou rna l o f Law and Econom i c s 2 1 1 ( 1 9 7 6 )

- 3 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

tru ly-a-phy s i c i a n

APPEN DIX A

MODEL OF PHYSICIAN STOCK AND INCOME USING DOMINANT FIRM -shyCOMPETITIVE FRINGE FRAMEWORK

Dema nd for a stock of phy s i c i ans in yea r t MDSt is

a s s ume d to be pos it i ve ly re l at e d to a ve ctor of dema nd s h i fters

Z inc lud i ng popu l at i on cha r a cte r i st i c s such as numbe r pe rsona l -t i ncome ins urance cove rage age d i st r i but i on and othe r hea lth

status det e rm i na nts as we l l as the pr i c e of subst itutes It is

negat i ve ly re l ated to the pr i ce of phy s i c i a ns the i r re a l

40e a rn i ngs MDINCt

In th i s s imp lest ca s e th en a l i ne a r ma r k et dema nd cu rve

f or a stock of phy s i c i a ns ca n be wr itten as

=( l a ) M DSt a + a MDINCt + a Zt -2-0 1

where abull is a row ve ctor w it h al l e leme nts gre ate r tha n ze ro -2 ( the eleme nts of Z the dema nd det e rmi na nt s have bee n so

-t def i ne d ) and i nd i cates a row ve ctor wh i l e denotes a

co lumn ve ctor

The ma rg i na l cost to the dom i na nt f i rm of supp l y i ng an an

a d d it i on a l phy s i c i a n can be vi ewed as the tota l opportun ity co st

( TOC ) of that ind i v i du a l not pu rsu i n g a ne xt be st ca re e r and can

be me a s u red as that i ncome at wh i c h he she is i nd i f f e re nt between

becom i ng a phy s i c i a n or pursu i ng the alternat i ve ca ree r The

40 Noeth e r sup r a note 1 d i s cu s ses the va l i d ity of tra n s l at i ng what is serv i ce dema nd i nto a dema nd for stock

- 40 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

shape of th i s cu rve de p e nd s on the qua l ity ( opportu n i ty cost )

p atte rn acros s the poo l of app l i c a nts to med i c a l schoo l S i nce

by most accounts the AMA-s a n ct i oned domi na nt f i rm has re str i cted

the supp ly of med i ca l schoo l spa c e s to the exte nt that ma ny

h i gh ly qua l i f i ed app l i cat i on s are re j e cted each yea r and s i nce

med i c a l stude nts rep re s e nt su c h a sma l l proport i on of al l co l l e ge

graduate s it seems re a s o n a b l y accu r ate to pos it the ma rg i na l

c ost cu rve as hor i z onta l i n the re l e va nt ra nge that is

( 2 a ) MCt = TOCt

where TOCt i s not a fu nct i on of the numb e r of phy s i c i a ns

The ma r g i na l cost cu rve ne ve rth e le s s sho u l d re f le ct any

e x i stent ba r r i e rs to entry du e for examp le to re str i ct i ons on

the numbe r of me d i c a l schoo l spac e s I f the ba r r i e r to ent ry is

tota l ly bi nd i n g the cost cu rve be come s ve rt i ca l when the con shy

stra i nt i s re a c he d

The compet it i ve f r i nge ( CNL ) has an upw a rd slop i ng supp ly

c u rve w ith re s p e ct to phy s i c i a n income ( M DINC )

( 3 a ) CNLt

where FTOC the f r i nge opportu n ity (ma r g i na l ) cost and R a

ve ctor of re str i ct i ons wh i ch the domi na nt f i rm impos es to impede

compet it i ve ent ry are paramet e r s det e rm i n i ng th e le ve l of the

s up p l y curve Thu s b gt 0 wh i l e b lt 0 and b lt 0 1 - 2 3

-4 1 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

-

In th i s ca s e the dom i nant f i rm faces a re s i du a l dema nd

cu rve equ a l i ng th e ma r k et dema nd ( l a ) less th e f r i nge sup p ly

( 3a )

The equ i l i br i um domi na nt f i rm stock oc c u r s where ma r g i n a l

revenue de r i ve d from th e re s i du a l dema nd curve equ a l s TOC The

tot a l sto ck equ a l s the sum of the dom i na nt f i rm bull s opt i mum le ve l

a nd the compet it i ve f r i nge or

( 4a )

MDSt =

b 3 + FTOCt + b1 MDINCt

S o l v i ng ( la ) for MDINC su bst itut i n g it into (4a ) and s imp l i f y shy

i ng y i e l d s th e re du c e d form

Ia

0 a

0b 1_ a 1 b

0 a a b 1 - 2 ( a -b ) + -4 TOCt + ( l shy

a l -b l ) l l

coe f f i c i e nt s on the z ve ctor are pos it i ve and thos e

MDS( 5a )

where th e

t =

on

TOC FTOC and th e R ve ctor are ne gat i ve

Due to unc e rta i nty about futu re demand andor co st cond i shy

t i ons pos s i b le s h i fts in the ma rket stru ctu re comb i ne d w ith

cost l y ad j u stme nt du e to the le ngth of tra i n i ng t i me requ i red to

be come a phy s i c i a n it is l i k e l y that the obs e rve d stock of

p hy s i c i a ns d i f f e r s f rom the equ i l i br i um va lue at any gi ven po i nt

i n t i me As in oth e r work study i ng the det e rm i n a nt s of an

-4 2shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

occupat i on bull s supp ly 4 1 we can pos it a part i a l ad j u stment mode l

where the cha nge i n actu a l stock from one year to the ne xt

repre s e nts on ly pa rt of the d i f f erence betw ee n th i s yea rbull s

e qu i l i br i um va lue and last yearbull s actu a l stoc k More ove r s i n ce

t ra i n i ng requ i re s fou r yea rs of med i ca l schoo l and one to f i ve

more of a graduat e program ( usua l ly f i ve ye a r s unt i l l i c e ns u re

f rom the beg i nn i ng of me d i ca l schoo l and th en one to fou r mo re of

res i de ncy tra i n i ng once l i ce n s ed ) th e p l a n n i ng hor izon is at

l e a st f i ve years long so that

( 6 a ) MOSt - MOSt - 1 = y ( MOSt l t - 5 - MOSt - l l t - 5 )

where MOSt = actu a l stock in yea r t

MOSt = actu a l equ i l i b r i um stock in year t

MOSt - l l t - 5 = pre d i ct ion made in t - 5 of actu a l stock that w i l l ex i st in t-1

MDSt l t - 5 = pre d i ct i on ma de in t - 5 of equ i l i br i um stock f or yea r t

y = ad j u stme nt rate to ch anges in equ i l i br i um p e r c e i ve d i n yea r t- 5

I f we ass ume that pre d i ct i ons from t- 5 ba s e d on ly on fore shy

c a sts of dema nd and cost va r i a b l e s of the opt ima l stock MOSt

a re unb i a s e d the n i s an unb i as e d est imator of MOStbullMDSt l t - 5

How mu ch unce rt a i nty e x i st s i n yea r t - 5 about the actu a l stock

4 1 B Pet e r Pas h i g i an The Ma rk et for Lawye rs The Determ i n a nt s o f the Dema nd a n d Supp l y of Lawye rs 20 Jou rna l of Law and E conom i c s 53 ( 1 9 7 7 )

- 4 3 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

e t - E ( l -d ) i - 1 ( FMG t - 1 - FMGt - i l t - 5 ) 4 3

that ex i s t s in yea r t - 1 I t has bee n shown that th i s unce r t a i nty

can be re duced to ignorance abou t the exa c t numb e r of FMG s

e n t e r i ng in the inte rven i ng fou r yea r per i od 4 2

I ncorpor a t i ng the s e s imp l i f i ca t i on s into ( 6a ) the ac t u a l

p hy s i c i an s t o c k ca n be e xp re s s ed as

F l ( 7a ) MDS t = yMDS t + ( 1 - y ) MDS t - 1 + yet + lit

F whe re et rep r e s e nt s unce rt a i nt y ove r FMG ent ry and is de f i ned as

F 4

i = l

whe re FMG = numb e r of fore i g n me d i ca l gra d u a t e s obt a i n i ng new us me d i c a l l i ce n s e s in yea r t

and d = yea r ly dep re c i a t i on ra t e of l i ce n s e d phy s i c i a ns and s tu d e nt s as s umed to be con s t ant ove r t ime

lu rep re s ents the ra ndom error i n pre d i c t i on s of the op t ima l t

s t ock 4 4

4 2 Noe th e r sup r a not e l

4 3 E ( e F eF ) i O is not ne ce s s a r i ly zero i f for examp l e t t - 1

gove rnme nt po l i c i e s w i t h re s p e c t to imm i g r a t i on are not pe r f e c t ly p re d i c t e d and are au tocorre l a t e d (more th a n one per i od long ) once

eFe n a c t e d More ge n e ra l ly c a n be thou ght of as the re su l t of t

any pre d i c t i on errors in yea r t for examp le une xpe c t e d va r i a shyt ions i n the dep r e c i a t i on ra t e may a l s o y i e l d fore c a s t error s

4 4 Actu a l ly i t cou l d be autocor re l a t e d for up to f i ve pe r i ods i f i n f o rma t i on abo u t pe rma ne nt changes in dema nd o r cos t is re l e a s ed between t - 5 and t

- 4 4 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

S ubs t i tu t i ng for MDS f rom ( 5a ) y i e l ds t

( 8 a ) MDS t = f ( a o

F + ( 1 - y ) MDS + ye t bullt - l

F i na l l y as sum i ng tha t phy s i c i a n earn i ng s ad j u s t to c le a r the

ma rket

I

( 9a ) M D I NC t

bullb l a b

+ [I2 ( 1 - --shya -b ) - 1 ] -

z_t + - 2 R

1 1 a l 2 ( a l -b l ) -t

Y2

The coe f f i c i e n t s on TOC and FTOC are po s i t i ve on MDS t - 1

a n d eF they are ne g a t i ve

- 4 5 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

3

u s annua l u s Bu reau of the Censu s HIStor i c a l Stat i s t i c s of the Un i t ed States f rom Co lon i a l T ime s to 1 9 7 0 ( he re a f t e r H i s tor i c a l S t a t i s t i c s ) a n d u s Pub l i c H e a l t h S e rv i c e Hea l t h Manpowe r Sou rce book

APPENDIX B

DESCRI P T I ON OF TH E DATA

Th i s app e nd i x de s c r i bes the dat a used i n es t i mat i ng the

mode l and dep i c t e d in the plo t s I t also c i t e s the i r sou rce s

1 MDS S tock of ac t i ve u s phy s i c i a ns

S ou rce s Ame r i c an Med i c a l As soc i at i on ( he re a f t e r AMA ) D i s t r i bu t i on o f Phy s i c i an s in the

v 20

Th e numbe r of act i ve MD s is ava i l a b l e on ly for s e l e c t e d yea rs pr i or to 19 6 3 ( 1 9 5 0 1 9 5 5 and 19 6 0 ) The oth e r yea r s f i gures have bee n es t i ma t e d by tak i n g 9 5 o f the tot a l numb e r o f phy s i c i a n s whe re 9 5 of the tot a l numbe r of phy s i c i a ns where 9 5 i s the a v e rage proport ion th at he l d for a l l yea rs from 1 9 5 0- 1 9 6 6 for wh i c h dat a we re ava i l a b l e

2 DNL New ly l i c e n s e d phy s i c i a n s tra i ne d i n dome s t i c sch oo ls Phy s i c i a n s who re ce i ve the i r i n i t i a l u s l i c e n s e ( a re add i t i on s to the me d i ca l prof e s s i on ) who gr aduated f rom an accre d i t ed u s or Canad i an s choo l

S ou rce AMA Med i ca l L i c e n s u re Stat i s t i c s annua l

FMG F ore i gn Med i c a l Gradu a t e s New ly U S - l i ce n s e d g ra d u a t e s o f fore i gn med i ca l schoo l s --they inc lude both fore i gn -born and us ci t i z e ns who re ce i ved the i r t ra i n i ng abroa d

Sou rce s e e DNL

( The above th ree va r i a b l e s are a l l de f l a t e d by the u s re s i dent p opu l a t i on i n th e es t ima t i on )

4 I NCM u s d i s p o s a b l e pe rsona l income i n 1 9 7 2 $

S ou rce H i s tor i ca l Stat i s t i c s and u s Bu reau of the Ce nsu s Sta t i s t i c a l Abs t ra c t of the U n i ted S t a t e s ( he re a f t e r St a t i s t i c a l Abs tra c t ) annu a l

- 4 6 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

Program

PUB I NS Percent age of popu l a t i on enro l l e d in pub l i c i ns u rance program pr i ma r i ly compos ed of Me d i c a re a nd Med i c a i d re c i p i e n t s

S ou r ce s U S Hou se of Rep re s e n t at i ve s Da t a on the Med i ca i d ( 1 9 7 7 ) and Hea l t h C a r e F i na n c i n g Admi n i s t ra t i on The Med i c a re and Med i ca i d Dat a Book 9 8 1 )

6 TOC Tot a l Oppor tun i ty Cos t --the ca l c u l a t ion of TOC is d e r i ve d e l s ewhere 4 5 Only its components are d e s cr i be d here

a C o l l e ge Gra duate Earn i ngs To be s t ho l d qua l i ty cons tant we look a t ea rn i ng s of ful l shyt ime work e rs w i t h fou r yea rs of co l le g e ove r a ge 2 5 bo th sexe s ( Me a n a f t e r tax 1 9 7 2 $ )

S ou rce s Current Popu la t i on Repor t s Ser i e s P - 6 0 us Bureau of the Censu s va r i ou s i s s u e s and H i s to r i c a l S ta t i s t i c s

b Tax Ra tes To de r i ve a f t e r tax income s ma r g i na l t a x ra t e s are ca l c u l a t ed f rom da t a on t ax payme n t s by income bra c k e t fou nd i n the I nt e rn a l Reve nue Se r v i c e pu b l i c a t i on S ta t i s t i cs of I ncome I nd i v i d u a l I ncome Tax Ret u rn s us Dep a rtme n t of the Tre a s ury annua l

c Lengt h of Me d i c a l Tra i n i ng T i me ca l c u l a t ed as f ou r yea rs of me d i c a l s choo l + g yea r s of g ra duate med i c a l t ra i n i ng - ( 4 -c ) yea r s le s s t h a n a fu l l fou r ye a r s o f co l l ege whe re

g t i s ca l c u l a t e d as 4x ( U S Res i de n t s amp I nt e rns ) t

3 E us M e d graduates t - i

i = O

c i s the ave rage numbe r of yea rs spe n t i n col lege pr i or t o med i c a l schoo l

S ee Noe t he r supra not e s 1 and 1 1

- 4 7 shy

4 5

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

br i n g i ng 1 9 7 9 4 7 are on ly

Sou rce A l l dat a requ i re d for ca l c u l a t i on of c and g are in annu a l Educa t i on

d D i s count ra t e 1 0 is used --the ra nge of d i s shycount ra t e s used i n mos t stud i e s is f rom 0 8 t o 1 2

e Net Tu i t i o n C a l c u l a t e d as ( Tot a l Tu i t i on P ayme n t s to Med i c a l Schoo l s le s s Tot a l S ch o l a rs h ip s g i ven to a l l Med i c a l stude n t s f rom a l l sources ) numbe r of stude n t s

S ou r ce Annu a l Educa t i o n numbe r of JAMA

numbe r of the Jou rn a l of the AMA ( he re a f t e r JAMA )

f I ncome ea rne d wh i le i n s choo l Fol low i ng B e c k e r 4 6 i t is as sume d th at med i c a l students earn 1 4 of what a co l le g e de greed work e r wou l d earn ( by work i ng summe rs andor part shyt ime ) Th i s as sumpt i on i s a l s o used for the r e s i de ncy tra i n i ng pe r i od Actu a l re s i dency s a l a r i e s have at lea s t s i nce 1 9 7 0 be e n con s i de ra b l y h i gh e r than 1 4 of the co l le g e wage bu t hou rs wor k ed ave r a ge ove r 7 0 pe r week thu s the hou r ly wage to about $ 4 2 2 i n Act u a l da t a on res i de ncy s t i p e nd s ava i l a b le annu a l ly be g i nn i ng i n the ea r l y 1 9 6 0 s it is a l s o di f f i cu l t to d i s ce rn how mu ch in i n -k i nd payme n t s has be e n g i ve n t o re s i de nt s

7 e F and e f 5 -year fore c a s t erro r s of FMG entry where

f =e t FMGP t - FMGP t l t - 5 and

4F f =e t ( 1 -d ) l - 1 e t - ii l

4 6 Gary Be c k e r Huma n Cap i t a l ( 2nd ed 1 9 7 5 ) at 1 5 3

Dou g l a s Houg h The Econom i c S t a t u s of Res i de n t Phy s i c i a n s Resu l t s f rom the Survey of Res i d e n t Phy s i c i a ns i n Prof i l e of M ed i c a l Pra ct i c e ( Ame r i ca n Med i c a l Assoc i at ion 1 9 8 1 ) at 8 3 - 9 6

- 4 8 shy

4 7

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy

( FMGP t i t - 5 ) moae l s

we re ca lcu shyl at e d from Box -Jenk i ns AR l ba s e d on the 1 5 y e a r s o f dat a end i ng 5 yea r s be f ore t h e pred i c shyt io n d is the FMG dep re c i a t i on ra t e as sume d t o b e a cons t a nt 2

The 5-yea r pre d i c t i on s

8 M D I NC Mea n us phy s i c i an i n come i n a f t e r -t a x 7 2 $ The ser i e s is sp l i ce d from the fou r se r i e s l i s t e d b e l ow s i nce no s i ng l e one prov i de s con s i s tent numbe rs for the ent i re pe r i od

S ource s AMA Prof i l e of Med i c a l Pra c t i c e and

i s s ue s

Tax da t a s e e TOC

9 CONS Cumu l a t i ve sum in bi l l i on s of 1 9 7 2 $ of con s t ruc shyt i on fu nds g i ve n to med i c a l schoo l s und e r federa l p rogr ams

Source us Hou se of Rep re s e nt a t i ve s Hea l t h M a npower Dat a Book ( 1 9 8 0 )

1 0 CAPS C ap i t at i on funds in bi l l i on s of 1 9 7 2 $ gi ve n to med i c a l s choo l s unde r fede r a l programs

S ou r c e S ee CON S

1 1 NOL I C We i gh t e d pe r c e n t age of s t a t e s that re f u s e to l i c e n s e F MG 1 s S ta t e s are we i gh t e d by the i r propor t i on of t o t a l nat i on a l pe r s on a l hea l t h care expe n d i tu re s a f t e r 19 6 5 and by pe r s on a l income be f ore 1 9 6 5

S ou rce s AMA Med i c a l L i ce n s u re S t a t i s t i cs an nua l

Ref e re nce Dat a on Prof i le of Me d i c a l Pract i ce va r i ou s i s s u e s M e d i c a l Econom i c s Company Med i ca l E conom i c s va r i ou s i s s u e s us Dep a rtme nt of Comme rce Survey of Current Bu s i ne s s Ju ly 1 9 5 1 us Dep a rtme nt of the Tre a s u ry I n t e rn a l Revenue Serv i ce Stat i s t i c s of Income I nd i v i du a l Income Ta x Re t u rn s va r i ou s

a n d H 1 s tor 1 ca l S t at 1 s t 1 c s and Stat i s t i ca l Abs t ra c t

1 2 C I T I Z We i gh ted pe r c e nt a g e of states c i t i z e n s h i p of FMG l i c e n s e e s s ou rce s see NOL I C

tha t F or

requ i re we i gh t s

us and

1 3 TNGTM We i gh t e d ave r a ge numbe r of yea r s ( re s i de n c y ) t ra i n i n g requ i re d of

w e i gh t s and sourc e s see NOL I C

of gra du ate FMG l i ce n s ee s F or

- 4 9 shy